A wind turbine aeroelastic response analysis method based on measured wind conditions and joint simulation

Abstract

The application discloses a wind turbine aeroelastic response analysis method based on measured wind conditions and joint simulation, and belongs to the technical field of wind power generation. The method comprises the following steps: obtaining measured wind conditions and fitting modeling of wind load; constructing a high-fidelity coupling model; constructing a measured wind load aeroelastic model; and performing parameterized analysis and application. The method realizes deep integration of regional measured wind conditions, a high-precision rigid-flexible coupling structure model and an aerodynamic load calculation model through joint simulation technology, overcomes the limitations of traditional methods using a simplified wind model and a rigid assumption, and greatly improves the accuracy and engineering practical value of aeroelastic response simulation of a large wind turbine in a complex real wind field.

Description

Technical Field

[0001] This invention belongs to the field of wind power generation technology, specifically relating to a method for analyzing the aeroelastic response of wind turbines based on measured wind conditions and co-simulation. Background Technology

[0002] As a crucial component of renewable energy, wind power generation relies heavily on the aeroelastic response of its core component—the wind turbine blade—under complex wind conditions. This response directly impacts the overall operational safety and lifespan of the turbine. Therefore, establishing high-fidelity aeroelastic simulation models has become a research hotspot for accurately assessing the dynamic characteristics of the blades in real wind environments.

[0003] In the existing technology, several aeroelastic modeling methods for wind turbines have been proposed. For example, Chinese invention patent CN117436322B discloses a wind turbine blade aeroelastic simulation method based on blade element theory. It achieves fluid-structure interaction by applying volume forces in the computational domain and iteratively solving the flow equations. However, its wind field is synthesized from parameters such as the wind profile power exponent, lacking direct input of measured wind conditions in specific regions, and relies on a self-developed flow field solver, limiting its engineering applicability. Another example is Chinese invention patent CN115544830B, which proposes a method for analyzing the dynamic response of wind turbines under random wind loads. It uses the Kaimal spectrum and harmonic superposition method to generate turbulent wind speed time histories, and applies the aerodynamic loads unidirectionally to the finite element model after calculating them using CFD. Although it can reflect the statistical characteristics of wind, it cannot capture unsteady and sudden wind events (such as gusts and wind shear) in real wind farms, and it does not achieve real-time feedback of structural deformation on aerodynamic loads. It is a unidirectional coupling and difficult to accurately simulate aeroelastic effects. In addition, Chinese patent application CN121118765A discloses a semi-coupled aeroelastic modeling method that considers nonlinear deformation. Although it constructs a high-precision geometrically accurate beam model for structural solution, the aerodynamic load still needs to be pre-exported from commercial software (such as Bladed). The coupling process is "offline" semi-coupled, which cannot realize the synchronous interaction between aerodynamics and structure in the time domain, thus limiting the accurate capture of transient response.

[0004] In summary, existing technologies generally have the following shortcomings:

[0005] (1) Wind condition models are mostly based on theoretical spectra or simplified assumptions, and lack effective use of measured wind data in specific regions;

[0006] (2) The coupling between aerodynamics and structure is mostly unidirectional or semi-coupled, and bidirectional real-time interaction has not been achieved;

[0007] (3) The modeling process relies on self-developed solvers or commercial software outputs, which lacks versatility and engineering applicability.

[0008] Therefore, there is an urgent need for a wind turbine blade aeroelastic simulation method that can integrate real wind conditions, support efficient two-way coupling, and is easy to implement in engineering. Summary of the Invention

[0009] To address the aforementioned technical problems, this invention provides a method for analyzing the aeroelastic response of wind turbines based on measured wind conditions and co-simulation. By integrating measured wind conditions, accurate aerodynamic models, high-fidelity rigid-flexible coupled multibody structure models, and co-simulation technology, this method achieves efficient and high-precision analysis of the aeroelastic response of wind turbines under real and complex wind loads.

[0010] The core of this invention lies in constructing a high-precision "measured wind-borne aeroelastic model", including:

[0011] Measured wind conditions acquisition and wind load fitting modeling: Based on the measured wind speed data of lidar in a specific region (such as the wind zone of Inner Mongolia), a four-component superposition method (constant wind, gust, gradual wind, and random wind) is used to perform high-precision fitting and construct an unsteady inflow wind load model that can reflect the fluctuation characteristics of real wind conditions.

[0012] High-fidelity coupled model construction: First, the super-element method in multibody dynamics is applied to discretize flexible components such as wind turbine blades and towers into multiple rigid bodies connected by hinges and spring dampers, establishing a rigid-flexible coupled multibody model of the whole machine considering structural flexibility. Second, a modified blade element momentum theory (BEM) is adopted and wind shear and tower shadow effects are introduced to establish an aerodynamic load calculation model.

[0013] Co-simulation integration (construction of measured wind-loaded aeroelastic co-simulation model): Through the co-simulation interface of multibody dynamics software and graphical modeling and simulation platform, the structural model and aerodynamic model are integrated to construct a measured wind-loaded aeroelastic model, realize real-time, bidirectional exchange of aerodynamic load and structural deformation data, and accurately simulate the aeroelastic coupling effect;

[0014] Parametric Analysis and Application: Using the established model, the aeroelastic response results under measured dynamic wind conditions are analyzed. Two key parameters in the measured wind load (peak gust and gust period) are systematically changed to analyze their impact on the aeroelastic response characteristics of the wind turbine, such as power output and blade tip displacement / velocity. This method can be used to evaluate the dynamic behavior of wind turbines under extreme wind conditions, providing a precise and reliable theoretical basis for optimizing high-durability blade structures and formulating safe operation control strategies.

[0015] This invention achieves deep integration of measured regional wind conditions, high-precision rigid-flexible coupled structural models, and aerodynamic models through joint simulation technology. It overcomes the limitations of traditional methods that use simplified wind models and rigid assumptions, and greatly improves the accuracy and engineering practical value of aeroelastic response simulation of large wind turbines in complex real wind fields.

[0016] To achieve the above objectives, the present invention adopts the following technical solution:

[0017] A method for analyzing the aeroelastic response of wind turbines based on measured wind conditions and co-simulation includes the following steps:

[0018] Step 1: Use a lidar wind measuring device to obtain the actual wind speed data of the target wind field at the hub height and different vertical height layers, and use the four-component superposition method to fit the measured wind speed data to establish a measured wind load model that includes constant wind, gust, gradual wind and random wind components.

[0019] Step 2: Based on the structural parameters of the target wind turbine, the super-element method in multibody system dynamics is applied to discretize the flexible components of the wind turbine into multiple rigid bodies connected by kinematic pairs and force elements, establish a rigid-flexible coupled multibody dynamics model of the whole machine, and perform modal analysis on the rigid-flexible coupled multibody dynamics model of the whole machine to verify its accuracy.

[0020] Step 3: Based on the blade element momentum theory, and with the introduction of wind shear effect and tower shadow effect for correction, establish the aerodynamic load calculation model of the wind turbine.

[0021] Step 4: Import the rigid-flexible coupled multibody dynamics model established in Step 2 into the control system simulation environment to generate a mechanical control module. Integrate the mechanical control module with the measured wind load model established in Step 1 and the aerodynamic load calculation model established in Step 3. Real-time bidirectional exchange of aerodynamic load and structural response data is achieved through the data interface.

[0022] Step 5: Run the measured wind load aeroelastic model to simulate the operating state of the wind turbine under steady-state conditions and unsteady inflow wind conditions described by the measured wind load model, and output the wind turbine aeroelastic response results.

[0023] Step 6: By changing specific wind condition parameters in the measured wind load model, simulate and analyze the nonlinear aeroelastic response characteristics of the wind turbine under different wind conditions.

[0024] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0025] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above method.

[0026] Beneficial effects:

[0027] 1. This invention is the first to use lidar measured wind speed data from typical wind resource areas such as Inner Mongolia as simulation input. It employs a four-component superposition method to perform high-precision fitting of average wind, gradual wind, gust wind and random wind, effectively restoring the real unsteady wind characteristics of a specific wind farm. This overcomes the "generalization distortion" problem caused by traditional methods relying on Kaimal spectrum or wind profile models, and significantly improves the representativeness of simulation results to actual operating conditions.

[0028] 2. This invention constructs a rigid-flexible coupled multibody dynamics model (using the super-element method) and utilizes a graphical modeling and simulation platform for real-time data exchange via the Adams / Controls interface, achieving bidirectional feedback between aerodynamic loads and structural deformation within each integration step. Compared to the unidirectional load transfer of CN115544830B or the semi-coupled method of CN121118765A, this application can more accurately capture nonlinear aeroelastic phenomena such as flutter and stall of blades under strong wind disturbances.

[0029] 3. This invention fully utilizes the collaborative capabilities of mature commercial multibody dynamics software (such as Adams) and graphical modeling and simulation platforms (Simulink), avoiding the development complexity of self-built flow field solvers. At the same time, it does not rely on the intermediate output of dedicated wind power software such as Bladed, forming a modular, portable, and easily integrated simulation framework. It is convenient to deploy quickly under different turbine models or control strategies and has good engineering promotion value.

[0030] 4. The dynamic response data obtained by this invention based on real wind conditions and a high-fidelity coupling model can provide a reliable basis for blade structure optimization, fatigue life assessment and active / passive control strategy development, which helps to improve the operational stability and power generation efficiency of wind turbines in complex wind environments.

[0031] In summary, this invention achieves significant breakthroughs over existing technologies in three dimensions: wind condition modeling, coupling mechanism, and engineering implementation. Attached Figure Description

[0032] Figure 1 This is a flowchart of a wind turbine aeroelastic response analysis method based on measured wind conditions and co-simulation according to the present invention.

[0033] Figure 2 This is a schematic diagram of the measured wind load model in this invention.

[0034] Figure 3 This is a schematic diagram of the measured wind speed and the fitted wind speed curve in this invention.

[0035] Figure 4 This is a schematic diagram of the aerodynamic parameters of the airfoil section in this invention.

[0036] Figure 5This is a schematic diagram of the aerodynamic load calculation model used in this invention.

[0037] Figure 6 This is a schematic diagram of the super-unit model in this invention.

[0038] Figure 7 This is a schematic diagram of the rigid-flexible coupling multibody structure model of the blades and the whole machine established in this invention.

[0039] Figure 8 This is a schematic diagram of the mechanical control module generated in the graphical modeling and simulation platform by combining the output file of multibody dynamics software in this invention.

[0040] Figure 9 This is the wind turbine measured wind load aeroelasticity joint simulation model established in this invention.

[0041] Figures 10a-10d This is a schematic diagram of the aeroelastic response results under steady-state (rated) operating conditions in embodiment S5 of the present invention; wherein, Figure 10a This is a curve showing the change in impeller speed over time. Figure 10b The curve shows the change of impeller power over time. Figure 10c The time-domain response of the leaf tip flapping direction displacement. Figure 10d The time-domain response of the blade tip oscillation direction displacement;

[0042] Figures 11a-11c This is a schematic diagram of the aeroelastic response results under measured dynamic wind conditions in embodiment S5 of the present invention; wherein, Figure 11a The curve shows the change in wind turbine power. Figure 11b The time-domain response of the blade tip displacement. Figure 11c The time-domain response of the blade tip velocity;

[0043] Figures 12a-12f This is a schematic diagram illustrating the influence of gust peak value variation on aeroelastic response in embodiment S6 of the present invention; wherein, Figure 12a This is the peak value variation curve of the gust. Figure 12b The curve shows the change in wind turbine power. Figure 12c The time-domain response of the blade tip displacement in the flapping direction. Figure 12d The time-domain response of the blade tip displacement in the oscillation direction. Figure 12e The time-domain response of the blade tip velocity in the direction of the flapping motion. Figure 12f The time-domain response of the blade tip velocity in the oscillation direction;

[0044] Figures 13a-13f This is a schematic diagram illustrating the influence of gust period variation on aeroelastic response in embodiment S6 of the present invention; wherein, Figure 13a This is the curve showing the periodic variation of gusts. Figure 13b The curve shows the change in wind turbine power. Figure 13c The time-domain response of the blade tip displacement in the flapping direction. Figure 13dThe time-domain response of the blade tip displacement in the oscillation direction. Figure 13e The time-domain response of the blade tip velocity in the direction of the flapping motion. Figure 13f The time-domain response of the blade tip velocity in the oscillation direction. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0046] like Figure 1 As shown, the present invention provides a method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation, comprising the following steps:

[0047] Step 1: Acquisition of measured wind data and wind load fitting modeling, including:

[0048] Long-term wind condition monitoring was conducted at the target wind farm using a lidar anemometer, acquiring time-series data on wind speed and direction at hub height and different altitude levels. After cleaning and validity verification of the measured wind speed data, a four-component superposition method was used for mathematical fitting to establish a measured wind load model incorporating constant wind, gusts, gradual wind, and random wind. The expression for this model is:

[0049] (1)

[0050] In the formula: This is the actual measured wind speed; The constant wind component is the average wind speed, i.e., the stable background wind. This is the gust component, where wind speed increases or decreases rapidly within a short period of time; As the wind component is gradually changing, the wind speed exhibits a slow changing trend. This represents a random wind component. The four components are ultimately superimposed to form the dynamic wind load input signal, including the gust component. and random wind components The parameters (such as peak value, period, amplitude, etc.) can be adjusted according to the analysis needs. The goodness of fit is evaluated by mean relative error (MRE), mean absolute error (MAE), and correlation coefficient (r) to ensure that the fitted model can accurately represent the measured wind conditions.

[0051] Step 2: Establishment of the rigid-flexible coupled multibody structure model of the wind turbine, including:

[0052] Based on the geometric and structural parameters of the target wind turbine, a rigid-flexible coupled multibody structural model was established using the super-element method in multibody system dynamics. This included: discretizing flexible components such as blades and towers into multiple rigid super-elements connected by universal joints, rotary hinges, and nonlinear spring dampers, accurately describing the elastic deformation of components with typically no more than 100 degrees of freedom (approximately 20-30 degrees of freedom per blade, 15-25 degrees of freedom per tower, and a total of 80-120 degrees of freedom for the entire model); treating components such as the hub and nacelle as rigid bodies; and connecting all components through constraints such as fixed hinges, rotary hinges, and ball joints to form a complete system multibody dynamics model. Modal analysis was performed on this model, and its natural frequencies and mode shapes were compared with authoritative data to verify the model's accuracy.

[0053] Step 3: Establishment of aerodynamic load calculation model, including:

[0054] A program for calculating aerodynamic loads was developed based on blade element momentum theory (BEM). This aerodynamic load calculation model considers tip loss and hub loss corrections, and introduces wind shearing and tower shadow effects to correct for incoming flow velocity, thereby improving the accuracy of aerodynamic load calculations. The inputs to this aerodynamic load calculation model include the hub height wind speed from S1. flapping velocities of the rigid body segments of the blade from S2 oscillation speed and aerodynamic twist and wind turbine speed and blade phase angle The output of this aerodynamic load calculation model is the axial force acting on the aerodynamic centers of each rigid body of the blade. Tangential force and aerodynamic torque The aerodynamic load calculation program is then encapsulated to form an aerodynamic load calculation model. Figure 5 ). Figure 5 The left side of the middle section shows the input parameters for each rigid body segment: V_HUI (flapping speed), V_BAI (swaying speed), Y (yaw angle), V_WIND (incoming wind speed), OMEGA (rotor speed), and sita (blade phase angle); the right side shows the output parameters for each rigid body segment: F_A (axial force), F_T (tangential force), M_O (aerodynamic torque), Alpha (angle of attack), F_T_0 (total axial thrust of the rotor), and power (output power).

[0055] , and The derivation process is as follows:

[0056] Velocity and geometry of blade element cross section:

[0057] like Figure 4As shown, the axial flow velocity at the wind turbine plane is Tangential incoming flow velocity The relative velocity encountered by the blade cross section can be obtained. :

[0058] (2)

[0059] In the formula: Wind speed; It is the axial induction factor; This refers to the rotational speed of the wind turbine; The position of the leaf element radius; Tangential induction factor; It is the sum of the local twist angle of the blade and the pitch angle; It is the angle of attack, which is the angle between the chord line in the airfoil and the direction of the incoming flow; It is the geometric inflow angle, i.e., the plane of rotation and the relative velocity. The angle between the three. The expression for the relationship between the three is as follows:

[0060] (3)

[0061] Furthermore, we can obtain:

[0062] (4)

[0063] In the formula: The waving speed of the aerodynamic center in the leaf element; denoted as ______, which is the oscillation velocity of the aerodynamic center in the leaf element.

[0064] Considering the aeroelastic coupling effect of flexible blade vibration when calculating aerodynamic loads, it can be seen from equation (4) that the flapping speed and oscillation speed affect the geometric inflow angle. This has an impact, and further affects the angle of attack. The size of the blades varies, and the local twist angle of each blade element also changes in real time, ultimately causing the aerodynamic load to change.

[0065] Aerodynamic calculations based on blade element theory: Projecting lift and drag onto the normal (axial) and tangential directions of the blade element:

[0066] (5)

[0067] (6)

[0068] In the formula: and These are the lift coefficient and drag coefficient, both of which are related to the angle of attack. The functions can be found by consulting the relevant manuals.

[0069] Acting on a length of Normal force on leaf element and tangential force Essentially, it is the integral of aerodynamic pressure in the corresponding direction. According to aerodynamics, the force acting on a minute element is proportional to the dynamic pressure. ,area And force coefficient. For those with A wind turbine with 12 blades has:

[0070] (7)

[0071] (8)

[0072] In the formula: air density, Let be the chord length of the airfoil.

[0073] Will and geometric relationships and Substituting into formulas (8) and (9), we get:

[0074] ;

[0075] ;

[0076] Substitute ,get:

[0077] (9)

[0078] Substitute and ,get:

[0079] (10)

[0080] aerodynamic torque It is tangential force Regarding the torque at the center of the wind turbine, i.e. From formula (10), we can obtain:

[0081] (11)

[0082] Momentum theory treats the wind turbine as a working disk and derives thrust (axial force) and torque by analyzing the momentum changes within the flow tube.

[0083] Thrust (axial force):

[0084] (12)

[0085] This is The momentum theory expression.

[0086] Torque:

[0087] (13)

[0088] The core idea of ​​blade element momentum theory is that, for the same blade element infinitesimal element, the aerodynamic load calculated by blade element theory must be equal to the aerodynamic load calculated by momentum theory. Based on this, equations are established and unknown induction factors are solved. and .

[0089] Let the thrust formula of the Yee element theory be equal to the thrust formula of the momentum theory:

[0090] ;

[0091] Simplify and introduce local realities The solution is:

[0092] (14)

[0093] Let the torque formula of the Ye element theory be equal to the torque formula of the momentum theory:

[0094] ;

[0095] Simplify and substitute local realities The solution is:

[0096] (15)

[0097] Using equations (2)-(15), combined with the lift-drag characteristic parameters of the airfoil, the required parameters are calculated using an iterative method. The iterative process is as follows:

[0098] (1) Factors and initialization;

[0099] (2) Calculate the inflow angle using equation (4) ;

[0100] (3) Calculate the local angle of attack using equation (3). ;

[0101] From local angle of attack Query its corresponding lift coefficient and drag coefficient ;

[0102] (5) Calculate using equations (5) and (6) and ;

[0103] (6) Calculate using equations (14) and (15) and ;

[0104] (7) Comparison and Compare the result with the previous calculation. If the custom iteration precision is met, the iteration ends; otherwise, the iteration continues.

[0105] (8) Calculate the local loads (axial force, tangential force, torque) on each part of the blade.

[0106] A Prandtl tip loss and hub loss model is introduced to correct for aerodynamic loads. The corrected axial force (thrust) is then calculated. Tangential force and torque Expressed as:

[0107] Axial force (thrust): ;

[0108] Tangential force: ;

[0109] Torque: ;

[0110] In the formula: This is the combined factor for tip and hub losses, with a value between 0 and 1. The overall load of the entire wind turbine is obtained by summing the integrals over all blade element infinitesimals.

[0111] Step 4, Joint Simulation Integration (Construction of a Joint Simulation Model of Measured Wind-borne Aeroelasticity), including:

[0112] The rigid-flexible coupled multibody structure model established in step 2 is exported to the simulation environment through its control system interface, generating an interactive rigid-flexible coupled multibody structure model mechanical control module (Adams_sub module), such as... Figure 8 As shown.

[0113] Simulation principle: Figure 8This demonstrates a co-simulation closed loop consisting of a Simulink control model and an ADAMS mechanical dynamics model. The entire process begins at the left (green) input interface. Nine control commands generated by the Simulink controller (such as the wind turbine speed Omega and the multi-channel motion speed command in_yp_III_b1) are synthesized into the bus signal outADAMS_uout by the Mux module, and sent together with the synchronization signals outADAMS_yout and outADAMS_tout to the middle ADAMS Plant module (red). This module embeds a high-precision mechanical system model (i.e., a rigid-flexible coupled multibody structure model) and calculates the dynamic response under the input in real time. Subsequently, dozens of physical quantities (such as acceleration a_FWJ and state feedback out_yp_III_b) are returned to Simulink from the right (blue) output interface. Some of these are immediately fed back to the controller to form closed-loop control, while others are recorded and output for subsequent performance analysis, plotting, and report generation, thus achieving a tight integration of control logic and high-precision physical simulation.

[0114] Module integration: In the graphical modeling and simulation platform, this module is integrated with the aerodynamic load calculation model encapsulated in step 3 and the measured wind load module established in step 1 through a data interface.

[0115] Data interaction: The calculation model of aerodynamic loads... , , As input, it is passed to the mechanical control module of the rigid-flexible coupled multibody structure model; the solution obtained by the mechanical control module is... , , State variables are used as feedback and passed to the aerodynamic load calculation model. This forms a closed loop of bidirectional real-time coupling between aerodynamics and structure.

[0116] Model building: A complete measured wind-borne aeroelastic joint simulation model is built in a graphical modeling and simulation platform. This model can simulate the dynamic response process of wind turbines under real wind conditions.

[0117] Step 5, Analysis and application of aeroelastic response characteristics, including:

[0118] Run the constructed measured wind load aeroelastic co-simulation model and set simulation parameters (such as wind speed, rotational speed, simulation time, and solver type). Change key parameters in the measured wind load model (e.g., peak gust). Gust cycle Parametric simulation studies were conducted to analyze the aeroelastic response characteristics of wind turbines under different wind conditions, including but not limited to:

[0119] Output performance: Wind turbine power fluctuation.

[0120] Structural response: The time-domain response of displacement and velocity in the blade tip flapping and oscillation directions.

[0121] Step 6: Optimize Design: Based on the analysis results, evaluate the operational safety of the wind turbine under complex wind conditions, identify dangerous operating conditions, and provide data support and theoretical basis for the structural optimization design of the blades and the formulation of control strategies.

[0122] Specifically, in step 1, real-time wind speed data of the target wind field is obtained by using a lidar wind measuring device to measure wind speed at the height of the wind turbine hub and at different vertical heights to obtain wind speed data with spatial distribution characteristics.

[0123] Specifically, in step 1, the gust component The mathematical expression is:

[0124] ;

[0125] In the formula, Peak gust value, m / s; The start time of the gust, in seconds; This refers to the gust cycle.

[0126] The gradual wind component The mathematical expression is:

[0127] ;

[0128] In the formula, The wind speed is for the gradual change of direction, in m / s; The peak value of the gradually changing wind is in m / s; The duration of the gradual wind change is in seconds. The start time of the gradual wind change is in seconds. The time for the gradual wind to end is s.

[0129] The random wind component The mathematical expression is:

[0130] ;

[0131] In the formula: The wind speed is random, in m / s; The wind peak value is random, in m / s; Let be a random variable uniformly distributed between -1 and 1; The average distance of wind speed fluctuation is typically taken as 0.5π to 2π rad / s. It is a random quantity that follows a uniform probability distribution between 0 and 2π.

[0132] Specifically, in step 2, the super-unit method is as follows: the wind turbine blades are discretized along the spanwise direction and the tower along the height direction into at least four super-units. Each super-unit consists of multiple rigid body segments connected by universal joints and rotary hinges, and spring dampers are used to characterize the bending and torsional stiffness and damping characteristics of the unit.

[0133] Specifically, in step 3, the input parameters of the aerodynamic load calculation model include at least: the hub height wind speed provided by the measured wind load model, the flapping speed, oscillation speed and aerodynamic twist angle of each segment of the blade as fed back by the rigid-flexible coupled multibody structure model; the output parameters are the axial force, tangential force and aerodynamic torque acting on each segment of the blade.

[0134] Specifically, in step 4, the data interface is implemented as follows:

[0135] The aerodynamic forces and torques calculated by the aerodynamic load calculation model are transmitted as input state variables to the mechanical control module to drive the rigid-flexible coupled multibody structure model to move.

[0136] The displacement, velocity, and angle parameters obtained from solving the rigid-flexible coupled multibody structure model are fed back as output state variables to the aerodynamic calculation load model for calculating the aerodynamic load at the next moment.

[0137] Specifically, step 5 includes: analyzing the aeroelastic response characteristics of the wind turbine under transient or alternating loads by individually changing the gust peak value or gust period; later, the nonlinear aeroelastic response characteristics of the wind turbine under combined changing wind conditions can be further analyzed by coupling and changing characteristic parameters such as the gust peak value, gust period, and random wind peak value.

[0138] Specifically, in step 6, the specific wind condition parameters include peak gust and gust period; the aeroelastic response characteristics include wind turbine output power, blade tip displacement, and blade tip velocity.

[0139] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. This embodiment takes a 5MW wind turbine as the analysis object, but the present invention is not limited to this specific model.

[0140] 1. Acquisition and Fitting Modeling of Measured Wind Condition Data:

[0141] At a wind farm in Inner Mongolia, a lidar anemometer was used to measure wind data at heights of 30m to 200m, centered at a hub height of 90m. Data from a representative spring period were selected, processed, and then fitted using a four-component superposition method. A constant wind speed was set. =6 m / s, gust component parameters are =19.8 m / s, =47.25s, =4.05s, random wind peak =1.0. The fitting results show MRE of 0.1366, MAE of 0.4597, and correlation coefficient r of 0.7833, which meet the requirements. Figure 2 As shown.

[0142] 2. Establishment of a rigid-flexible coupled multibody structure model for wind turbines:

[0143] like Figure 6 As shown, a 5MW wind turbine 3D model was created in SolidWorks. Each blade was discretized into 13 rigid bodies, and imported into multibody dynamics software (such as Adams) using the superelement method (every 4 rigid bodies form a superelement). Universal joints, rotary hinges, and spring dampers were then added. The tower was processed in the same way. The entire turbine model consists of 56 rigid bodies. Figure 7 As shown, superunit I, superunit II, superunit III, and superunit IV are formed and discretized into 13 rigid bodies, respectively. - Modal analysis was performed, and the results showed that the simulated values ​​of the first-order flapping frequency, first-order oscillation frequency, and second-order flapping frequency of the blade were 0.6763 Hz, 1.1011 Hz, and 1.8824 Hz, respectively. The errors compared with the corresponding reference values ​​of 0.6686 Hz, 1.0793 Hz, and 1.8558 Hz for the blade of a 5 MW wind turbine were 1.15%, 2.52%, and 1.43%, respectively, indicating that the model verification was successful.

[0144] 3. Establishment of aerodynamic load calculation model:

[0145] Write an aerodynamic load calculation program in MATLAB based on BEM theory, including corrections for tip loss, hub loss, wind shear, and tower shadow effect. Utilize airfoil aerodynamic data (C) for a 5MW blade. l C d ).

[0146] 4. Co-simulation integration (construction of a co-simulation model of wind-borne aeroelasticity based on actual measurements):

[0147] The rigid-flexible coupled multibody structure model from section 2 was exported as a module recognizable by the graphical modeling and simulation platform. The model, as shown in the abstract, was then built within this platform. Three aerodynamic modules (green boxes) were set, each corresponding to one of the three blades (with a phase angle difference of 120°). The red box represents the measured wind load fitting model; the purple box represents the mechanical control module of the rigid-flexible coupled multibody structure model of the wind turbine; and the blue box represents the data acquisition module. The initial simulation time was set to 90 seconds, with the rotor speed linearly increasing from 0 rpm to the rated speed of 12.1 rpm for the first 30 seconds, and then remaining constant. The solver was ode45, and the simulation time and solution step size were then varied for the solution.

[0148] 5. Analysis and application of aeroelastic response characteristics:

[0149] First, the model was run at the rated wind speed (11.4 m / s steady wind) to verify its correctness: the rotor speed was set to slowly increase from 0 r / min to the rated speed of 12.1 r / min (1.267 rad / s) within 30 s. Considering wind shear and tower shadow effects, ode 45 was selected as the numerical integrator, with a time step of 0.005 s and a simulation time of 80 s. The rated power was approximately 5.1 MW and the blade tip flapping displacement was approximately 5.4 m. The errors compared with the literature results (rated power 5.2 MW, blade tip flapping displacement 5.56 m) were 2% and 2.88%, respectively, both less than 5%. This demonstrates the reliability of the established measured wind-borne aeroelastic model.

[0150] Figures 10a-10d The results represent the aeroelastic response under steady-state (rated) operating conditions. Figure 10a The impeller speed changes over time, and the speed remains constant after reaching the rated speed in 30 seconds. Figure 10b The curve of rotor power versus time shows that the power fluctuation of the wind turbine tends to stabilize after reaching the rated speed, fluctuating around 5.1 MW. Figures 10c-10d The time-domain displacement response curves of the blade tip flapping and flaring directions show that, after reaching the rated speed, the blade tip flapping displacement fluctuates between 4.1 and 5.5 m, with an average displacement of approximately 4.63 m. The blade tip flaring displacement is relatively smaller, fluctuating between -0.1 and 0.8 m, with an average displacement of approximately 0.42 m. This is because the blade displacement in the flapping direction is due to the axial component of the aerodynamic force, thus the flapping displacement fluctuates with the axial force. In contrast, the flaring displacement occurs due to the greater blade stiffness and the gravitational load acting on the blade.

[0151] Then, the measured wind load model established in step 1 is used as input to analyze the response under its action. Figures 11a-11c This represents the aeroelastic response results under measured dynamic wind conditions. Figures 11a-11bIt can be seen that the changing trends of wind turbine power, blade tip flapping displacement, and blade tip oscillation displacement are related to... Figure 3 The wind speed change trend shown is consistent, and the wind turbine is in normal operation. When the system is hit by gusts, the power output increases sharply, and the displacement of the blade tip flapping and oscillation direction also changes significantly. After the gusts pass, due to the structural rigidity of the wind turbine blades, the fluctuations in the rotor power and blade tip displacement converge to a stable fluctuation state.

[0152] Next, parameterization studies will be conducted:

[0153] (1) Change the peak value of gusts

[0154] The measured wind speed curve shows that the gust started at 47.2s and stopped at 51.3s, lasting a total of 4.1s. When the gust occurred, the wind speed at the hub height increased from 6.6m / s to 19.8m / s within 3s, and then decreased from 19.8m / s to 6.8m / s within 1.1s. Therefore, by changing the peak gust speed from the original measured 19.8m / s to 15.8m / s and then increasing it to 25m / s (cutoff speed), the changes in the peak gust speed are as follows: Figure 12a The aeroelastic response characteristics of wind turbines under different peak gusts were analyzed, and the results are as follows: Figures 12b-12f As shown.

[0155] Figure 12b The wind turbine power variation curve shows that gusts cause a sudden change in wind turbine power: a measured gust of 19.8 m / s peak power causes the wind turbine power to suddenly increase to 5.58 × 10⁻⁶ m / s. 6 After the gust ended, the power returned to normal. Due to the short duration, although the gust caused the power to briefly exceed the rated power, the increase was small and did not affect the normal power generation of the wind turbine. When the peak gust speed was reduced to 15.8 m / s, since it was close to the original wind speed, the power increased suddenly to 2.92 × 10⁻⁶ m / s. 6 W, but the change is small and within the range that the wind turbine blade structure can withstand, so it does not affect the normal operation of the wind turbine; when the peak gust increases to 25 m / s, the power change is unusually obvious, and the maximum power increases to 1.06 × 10 7 W reached 112% of the rated power, and after the gust ended, the power change was relatively large compared to the measured gust of 19.8 m / s and the lower peak gust of 15.8 m / s.

[0156] In addition, strong gusts of wind also cause the blades to deform dramatically. Figures 12c-12d as well as Figures 12e-12fThe time-domain responses of blade tip displacement and blade tip velocity are shown in the two directions of flapping and swaying, respectively. It can be seen that when the measured peak gust is 19.8 m / s, the blade tip flapping displacement is 5.72 m and the swaying displacement is 0.937 m; although the blade tip flapping velocity fluctuates significantly, it is within a reasonable range and the fluctuation returns to normal after the gust ends; the blade tip swaying velocity fluctuates normally without any obvious abrupt change.

[0157] When the measured peak gust speed was reduced to 15.8 m / s, the blade tip flapping displacement was 2.48 m and the oscillation displacement was 1.02 m. Although the amplitude of the blade tip flapping and oscillation displacement fluctuated slightly under the action of the gust, it tended to stabilize after the gust ended. This indicates that under the action of low peak gust speed, due to the stiffness of the blade itself, the fluctuation of the blade tip flapping speed and oscillation speed did not change significantly.

[0158] When the measured peak gust speed was increased to 25 m / s, the blade tip flapping displacement was 7.49 m and the flapping displacement was 2.29 m. Compared to the original peak gust speed of 19.8 m / s, the blade tip flapping displacement increased by 31% and the blade tip flapping displacement increased by 144%. At this point, the blade tip velocity fluctuated significantly in both the flapping and flapping directions, especially in the flapping direction. The fluctuation continued to increase even after the gust ended. This resulted in large deformation of the wind turbine blades, leading to blade structural damage. It is evident that during wind turbine operation, the effect of gusts on blade tip deformation in the flapping direction is greater than its effect in the flapping direction.

[0159] In summary, the curves showing the changes in rotor power, blade tip displacement, and blade tip velocity after the change in peak gust speed demonstrate that the measured change in peak gust speed has a significant impact on the aeroelastic response of the wind turbine. As the peak gust speed increases, the wind speed also increases, and the transient load on the wind turbine increases accordingly, leading to severe vibrations in the turbine blades. If the rotor speed is not adjusted or pitch control is not implemented in a timely manner, it will pose a significant threat to the safe and stable operation of the wind turbine.

[0160] (2) Change the gust cycle

[0161] Depend on Figure 3 The measured wind speed curves show that the gusts begin to act at 47.25s and cease at 51.3s, constituting one gust cycle. The effects of measured gust cycle variations on the aeroelastic response of the wind turbine were analyzed using 3-cycle and 6-cycle variations. Specifically, the gust cycle was increased from the original 1 cycle to 3 and 6 cycles respectively. The gust cycle variation curves are shown below. Figure 13a By changing the measured gust period while keeping other simulation conditions unchanged, the aeroelastic response characteristics of the wind turbine under different gust periods were analyzed. The results are as follows: Figures 13b-13f As shown.

[0162] Figure 12bThe curves showing the variation of wind turbine power over time under different gust cycles reveal that when the wind turbine is subjected to a single-cycle gust (i.e., the measured wind load), although there are sudden power fluctuations due to the short-duration impact of the gusts and low wind speeds, the power fluctuations tend to stabilize after the gusts end, having little impact on the normal operation of the wind turbine. When the wind turbine is subjected to three-cycle gusts, after being impacted by a brief period of continuous gusts, the power fluctuations are only significant during the first cycle of gusts, and the fluctuations tend to stabilize in the subsequent two cycles. After the three-cycle gusts end, the wind turbine power returns to normal fluctuations. When the wind turbine is subjected to six-cycle gusts (i.e., under continuous multi-cycle gust impacts), although there are sudden power fluctuations in the first four cycles, the fluctuations are relatively stable. When impacted by the fifth cycle of gusts, the wind turbine power experiences a sudden increase, especially during the sixth cycle, where the power increase is exceptionally significant, with the maximum power increasing to 1.38 × 10⁻⁶. 7 When the wind turbine reaches 176% of its rated power, the turbine blades are damaged under the alternating loads generated by multiple gusts of wind.

[0163] Figure 13c and Figure 13d The figures show the curves of blade tip displacement and blade tip velocity in the flapping and swaying directions as a function of time after changing the gust cycle. It can be seen that when the wind turbine is under a gust of one cycle, the change in blade tip flapping displacement follows the same trend as the wind speed, while the fluctuation amplitude of blade tip swaying displacement is relatively small, with a maximum displacement of 1m; the change trends of blade tip velocity in both flapping and swaying directions also follow the same trend as the wind speed.

[0164] When the wind turbine is subjected to three gust cycles, the blade tip flapping displacement fluctuates significantly under the impact of the gusts, returning to a stable state after the gusts end. The fluctuation amplitude of the blade tip oscillation displacement increases slightly, with a maximum displacement of 1.7m. After the gusts end, the fluctuation amplitude decreases and tends to stabilize. Although the blade tip velocity fluctuation varies, the fluctuation amplitude gradually stabilizes after the gusts end, although it is slightly larger than that of a single gust cycle. The reason for this is that the structural rigidity of the wind turbine blades can withstand the impact of short-term continuous gusts. After the gust cycle ends, the wind turbine blades still experience brief fluctuations due to the vibration not dissipating immediately. However, as the operating time progresses, they gradually return to normal operating conditions, without affecting the safety of the wind turbine.

[0165] When the wind turbine was subjected to six gust cycles, both the blade tip flapping and tip flaring displacements fluctuated significantly. The maximum deformation of the blade tip flapping displacement was 8.61m, and the maximum deformation of the tip flaring displacement reached 2.3m, with the fluctuations showing an increasing trend. Correspondingly, the blade tip velocity fluctuated significantly during the fifth gust cycle, and even after the gust ended, the blade tip velocity fluctuations remained severe and showed an increasing trend. The reason for this is that under continuous fluctuating loads, the blade structure vibrates, and the vibration frequency changes and increases continuously under continuous gust cycles, leading to irregular deformation of the wind turbine blades and ultimately damage.

[0166] In summary, based on the curves showing the changes in rotor power, blade tip displacement, and blade tip velocity after altering the gust period, it is evident that the measured periodic changes of the gust have a significant impact on the aeroelastic response of the wind turbine. If subjected to short-term gusts with low wind speeds, the wind turbine's output is not significantly affected, and it can operate normally. However, as the gust period lengthens, the gust fluctuation time increases, leading to a corresponding increase in the time it takes for the load on the wind turbine to change. With the continuous increase in the duration of the gust's effect, the fluctuations in various parameters of the wind turbine show an increasing trend, especially at the blade tips. This increasingly pronounced change makes the structural stiffness of the wind turbine blades unable to withstand the impact of long-term gusts, causing them to become highly susceptible to damage and severely impacting the safe operation of the wind turbine.

[0167] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0168] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above method.

[0169] In summary, the method of this invention can effectively and accurately analyze the aeroelastic response of wind turbines under real and complex wind conditions, and has important engineering application value.

Claims

1. A method for analyzing the aeroelastic response of wind turbines based on measured wind conditions and co-simulation, characterized in that, Includes the following steps: Step 1: Use a lidar wind measuring device to obtain the actual wind speed data of the target wind field at the hub height and different vertical height layers, and use the four-component superposition method to fit the measured wind speed data to establish a measured wind load model that includes constant wind, gust, gradual wind and random wind components. Long-term wind condition monitoring was conducted at the target wind farm using a lidar anemometer to acquire time-series data on wind speed and direction at hub height and different altitude levels. After cleaning and validity verification of the measured wind speed data, a four-component superposition method was used for mathematical fitting to establish a measured wind load model that includes constant wind, gusts, gradual wind, and random wind. The expression for this model is as follows: （1） In the formula: This is the actual measured wind speed; The constant wind component is the average wind speed, i.e., the stable background wind. This is the gust component, where wind speed increases or decreases rapidly within a short period of time; As the wind component is gradually changing, the wind speed exhibits a slow changing trend. The four components are random wind components; they are ultimately superimposed to form the dynamic wind load input signal, and the gust component is also included. and random wind components The parameters are adjusted according to the analysis needs; the goodness of fit is evaluated by mean relative error (MRE), mean absolute error (MAE), and correlation coefficient (r) to ensure that the fitted model can accurately represent the measured wind conditions. Step 2: Based on the structural parameters of the target wind turbine, the super-element method in multibody system dynamics is applied to discretize the flexible components of the wind turbine into multiple rigid bodies connected by kinematic pairs and force elements, establish a rigid-flexible coupled multibody dynamics model of the whole machine, and perform modal analysis on the rigid-flexible coupled multibody dynamics model of the whole machine to verify its accuracy. Based on the geometric and structural parameters of the target wind turbine, a rigid-flexible coupled multibody structural model is established using the super-element method in multibody system dynamics. This includes: discretizing flexible components such as blades and towers into multiple rigid super-elements connected by universal joints, rotary hinges, and nonlinear spring dampers, accurately describing the elastic deformation of the components with no more than 100 degrees of freedom; treating components such as hubs and nacelles as rigid bodies; and connecting all components through constraints such as fixed hinges, rotary hinges, and ball joints to form a complete system multibody dynamics model. Step 3: Based on the blade element momentum theory, and with the introduction of wind shear effect and tower shadow effect for correction, establish the aerodynamic load calculation model of the wind turbine. Step 4: Import the rigid-flexible coupled multibody dynamics model established in Step 2 into the control system simulation environment to generate a mechanical control module. Integrate the mechanical control module with the measured wind load model established in Step 1 and the aerodynamic load calculation model established in Step 3. Real-time bidirectional exchange of aerodynamic load and structural response data is achieved through the data interface. Step 5: Run the measured wind load aeroelasticity co-simulation model to simulate the operating state of the wind turbine under steady-state conditions and unsteady inflow wind conditions described by the measured wind load model, and output the wind turbine aeroelasticity response results. Step 6: By changing specific wind condition parameters in the measured wind load model, simulate and analyze the nonlinear aeroelastic response characteristics of the wind turbine under different wind conditions.

2. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 1, the measured wind load model constructed by the four-component superposition method is used to characterize the real unsteady wind characteristics of the target wind farm, wherein the parameters of the gust component and the random wind component are adjustable.

3. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 2, the flexible component includes blades and towers, the kinematic pair includes universal joints and rotary hinges, and the force element includes a spring damper, used to characterize bending and torsional stiffness and damping characteristics.

4. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 3, the inputs to the aerodynamic load calculation model include the hub height wind speed provided by the measured wind load model, and the flapping speed, oscillation speed and aerodynamic twist angle of each segment of the blade rigid body fed back by the rigid-flexible coupled multibody dynamics model.

5. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 4, the data interface is configured to: transmit the aerodynamic force and torque output by the aerodynamic load calculation model as input to the mechanical control module, and simultaneously feed back the displacement, velocity and angle parameters obtained by the mechanical control module to the aerodynamic load calculation model.

6. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 5, the aeroelastic response results include wind turbine power, blade tip flapping direction displacement, blade tip oscillation direction displacement, blade tip flapping direction velocity, and blade tip oscillation direction velocity.

7. The method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation as described in claim 1, characterized in that, In step 6, the specific wind condition parameters include peak gust and gust period. By changing these parameters, their influence on the aeroelastic response characteristics of the wind turbine is analyzed.

8. A method for analyzing the aeroelastic response of a wind turbine based on measured wind conditions and co-simulation, as described in any one of claims 1 to 7, characterized in that, The co-simulation environment enables bidirectional real-time interaction between aerodynamic loads and structural deformation within each integration step, forming a closed-loop coupling cycle.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of a wind turbine aeroelastic response analysis method based on measured wind conditions and co-simulation as described in any one of claims 1-8.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the wind turbine aeroelastic response analysis method based on measured wind conditions and co-simulation as described in any one of claims 1-8.

Images