Simulation system and simulation method for dynamic wake of wind farm based on digital twinning

Abstract

The application discloses a simulation system and simulation method for dynamic wake of a wind farm based on digital twinning, which realizes full-process closed-loop testing from a turbulent wind field to power generation by combining numerical simulation and physical simulation, and comprises a wind farm numerical simulation module, a wind turbine simulation module and a data visualization module; the wind farm numerical simulation module comprises a wind farm inflow turbulent wind speed generation unit, a wind farm wake model calculation unit and a wind turbine transmission chain rotating speed calculation unit, the wind turbine simulation module comprises a wind turbine transmission chain six-degree-of-freedom load test platform, and the data visualization module is used for displaying the effect of digital twinning; and the application can quantitatively evaluate the precision difference of different wake models on power prediction of downstream units, reveals the electromechanical response characteristics of the transmission chain under the excitation of dynamic wind speed, and provides an efficient test solution for wind farm optimization layout, wake control strategy design and transmission chain reliability research.

Description

Technical Field

[0001] This invention relates to the field of wind turbine generator and wind farm simulation technology, specifically to a simulation system and method for dynamic wake of a wind farm based on digital twins. Background Technology

[0002] A wind farm typically consists of dozens or hundreds of wind turbines arranged in an array. The dynamic response characteristics of each wind turbine under random wind excitation, as well as the flow field influence between turbines arranged in front and behind, can be considered as the dynamic characteristics of the wind farm. Among these, the wake effect during wind farm operation is one of the key factors affecting turbine output and power generation efficiency. The wake of upstream wind turbines significantly reduces the inflow wind speed of downstream wind turbines, causing power loss and exacerbating transmission chain load fluctuations. Simulating the dynamic characteristics of wind turbines and wind farms in a ground environment helps deepen the understanding of wind farm operation, facilitates the inspection and testing of wind turbines, and provides reference and basis for design improvements of wind turbines and wind farms.

[0003] Currently, existing methods generally simulate the transmission chain characteristics of wind turbines by using an electric motor to drive a generator, which cannot reproduce the wake effect of wind farms in a ground environment. Existing wind farm wake simulation methods mainly rely on numerical simulation and lack verification methods for physical transmission chains. Patent application (publication number: CN117592388A) discloses a CFD-based multi-machine wake simulation method for wind farms, which only uses numerical calculations to simulate wind farm wakes; patent (publication number: CN113656973B) provides a hybrid wake simulation method, system, device, and medium for wind farms, solving the problem that existing wind farm wake simulation methods cannot balance computational accuracy and computational speed, but this patent also only provides numerical simulation algorithms; patent application (publication number: CN119830521A) discloses a wake simulation method for a wind turbine motor experimental platform based on an analytical wake model, but this method only involves the static model of wind farm wakes and does not consider the dynamic characteristics of the wind turbine transmission chain. Summary of the Invention

[0004] The purpose of this invention is to address the problems existing in the prior art by providing a simulation system and method for dynamic wake of wind farms based on digital twins.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Firstly, a simulation system for dynamic wake of a wind farm based on digital twins is provided, including a wind farm numerical simulation module, a wind turbine simulation module, and a data visualization module. The wind farm numerical simulation module includes a wind farm inflow turbulence wind speed generation unit and a dynamic model coupling the wind farm wake with the wind turbine transmission chain. The dynamic model is used to output numerical simulation calculation results and issue speed commands. The wind turbine simulation module includes multiple wind turbine load testing platforms. The wind turbine load testing platforms use the speed commands as reference speeds for tracking and output test data. The data visualization module calls the wind farm numerical simulation module and receives the test data. The data visualization module is used to coordinate the wind farm numerical simulation module and the wind turbine simulation module to demonstrate the effect of digital twins.

[0006] This simulation system combines numerical and physical simulations to achieve closed-loop testing of the entire process from turbulent wind farms to power generation. It can quantitatively evaluate the differences in the accuracy of different wake models in predicting downstream unit power, reveal the electromechanical response characteristics of the drive train under dynamic wind speed excitation, and provide an efficient testing solution for wind farm optimization layout, wake control strategy design, and drive train reliability research.

[0007] This digital twin-based wind farm dynamic wake simulation system can not only generate wind farm wake results calculated by numerical simulation, but also conduct data verification and comparative analysis through the wind turbine load testing platform, taking into account both numerical calculation and measured results, and has good reliability.

[0008] Furthermore, the wind farm inflow turbulent wind speed generation unit has multiple built-in turbulent wind speed generation models, and different turbulent inflow wind speeds can be obtained by calling the turbulent wind speed generation models.

[0009] Furthermore, the dynamic model includes a wind farm wake model calculation unit and a wind turbine drivetrain speed calculation unit. The wind farm wake model calculation unit has multiple wind farm wake calculation models built-in, and different wind turbine wake fields are obtained by calling the wind farm wake calculation models. The wind turbine drivetrain speed calculation unit has wind turbine aerodynamic models, electromechanical control equations and operating modes built-in, and is used to calculate the drivetrain speed of multiple wind turbines to obtain the speed command.

[0010] Furthermore, the wind turbine load test platform includes a wind turbine nacelle under test and a non-contact electromagnetic force loading device connected to the wind turbine nacelle under test. The non-contact electromagnetic force loading device is connected to a drive motor through a permanent magnet eddy current coupler. The drive motor is used to load wind torque load and simulate the low-speed operation of the wind turbine.

[0011] Furthermore, the non-contact electromagnetic force loading device includes a housing and a flywheel inertia disk suspended within the housing. The input shaft of the flywheel inertia disk is connected to the permanent magnet eddy current coupler, and the output shaft of the flywheel inertia disk is connected to the wind turbine nacelle under test. The outer periphery of the housing is provided with a plurality of axially and radially arranged electromagnet assemblies. Each electromagnet assembly corresponds non-contactly to the axial or radial disk surface of the flywheel inertia disk. The combined attraction of the plurality of electromagnet assemblies acts on the multi-degree-of-freedom non-torque load of the flywheel inertia disk.

[0012] Secondly, a simulation method for dynamic wake of a wind farm based on digital twins is provided, the simulation method comprising the following steps: Construct the inflow turbulent wind speed generation unit for the wind farm, and call the turbulent wind speed generation model to obtain the turbulent inflow wind speed; A wind farm wake model calculation unit is constructed to calculate the wake wind speed based on the turbulent inflow wind speed. A wind turbine drive train speed calculation unit is constructed, and the dynamic model is obtained by coupling the wind turbine drive train speed calculation unit and the wind farm wake model calculation unit. Based on the dynamic model, the wind turbine main shaft speed and the wake wind speed, the aerodynamic torque of the wind turbine is obtained, the wind turbine drive train speed is calculated, and the speed command is sent to the wind turbine load test platform. The wind turbine load test platform simulates the low-speed operation condition of the wind turbine according to the speed command and outputs the test data. The data visualization module compares and verifies the received test data with the numerical simulation results output by the dynamic model, and corrects the numerical simulation results accordingly.

[0013] Furthermore, the turbulent inflow wind speed includes the average wind speed and the turbulent wind speed with random characteristics. The turbulent wind speed with random characteristics uses the power spectrum to define the density function and spatial coherence in the frequency domain, and then uses the inverse Fourier transform to generate a time series, so as to obtain the simulated turbulent wind speed of all simulation points in space at each time step.

[0014] Furthermore, based on the aforementioned turbulent wind speed generation model, an aerodynamic model of the wind turbine is established using blade element momentum theory, and the transmission chain equation of the wind turbine is established using a dual-mass block model, coupling the aerodynamic model and the transmission chain equation. On this basis, considering the wake effect of the wind farm and the lag or propulsion time of the incoming wind reaching subsequent wind turbines, a dynamic model is established based on the coupling of the wind farm, wind turbine aerodynamics, and transmission chain dynamics. The following model describes the dynamics of the transmission chain of each wind turbine within the wind farm and the wake propulsion process: , In the formula, Wind turbine jThe aerodynamic torque, spindle speed and front wake velocity, t 1, t 2, …, t j , …, t N Let T be the time required for different turbulent inflow wind speeds to reach the wind turbine sequentially, where T is the current time, and f and g represent functions respectively.

[0015] Furthermore, the wake calculation model includes the Howland, Gebraad, or Frandsen wake model, and the output test data includes parameter data such as output speed, voltage, and electric power.

[0016] Furthermore, the issuance of the speed command and the feedback of the test data are implemented using a communication protocol to encapsulate the command into a hexadecimal command frame, and the command is sent to the wind turbine simulation module in a cyclic manner at time step intervals through the serial communication module.

[0017] Compared with existing technologies, the beneficial effects of this invention are: 1. This simulation system achieves a closed-loop test of the entire process from turbulent wind farm to power generation by combining numerical simulation and physical simulation. It can quantitatively evaluate the accuracy differences of different wake models in predicting downstream turbine power, reveal the electromechanical response characteristics of the drive train under dynamic wind speed excitation, and provide an efficient testing solution for wind farm optimization layout, wake control strategy design, and drive train reliability research; 2. This invention adopts a digital twin approach, combining the wind farm wake model with the wind turbine drive train six-degree-of-freedom load test platform to form a closed loop, which can more realistically and accurately simulate the dynamic wake characteristics of wind turbines; 3. This digital twin form of wind power... The wind turbine wake simulation system not only generates numerical simulation results of wind farm wakes, but also enables data verification and comparative analysis through the wind turbine load testing platform, taking into account both numerical calculations and measured results, thus exhibiting good reliability. 4. The wind turbine load testing platform adopts a six-degree-of-freedom load testing platform solution. This platform not only simulates flywheel inertia but also uses a non-contact electromagnetic force loading method, featuring non-contact operation and no lubrication required, eliminating reliability and lifespan issues caused by contact fatigue and wear. Simultaneously, the flywheel inertia disk is levitated to the drive motor via a permanent magnet eddy current coupler, realistically reproducing the actual operating state of the wind turbine, thus exhibiting good accuracy. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall system for simulating the dynamic wake of a wind farm based on digital twins according to the present invention. Figure 2 This is a schematic diagram of the overall structure of the wind turbine load testing platform of the present invention; Figure 3 This is a schematic diagram of the non-contact electromagnetic force loading device of the wind turbine load testing platform of the present invention; Figure 4 This is a schematic diagram illustrating the establishment of the turbulent inflow wind speed according to the present invention; Figure 5 This invention provides a calculation model for the wind farm wake propulsion time. In the figure: 1. Wind turbine nacelle under test; 2. Non-contact electromagnetic force loading device; 201. Housing; 3. Permanent magnet eddy current coupler; 4. Drive motor; 5. Flywheel inertia disk; 6. Electromagnet assembly; 601. Electromagnet body; 602. Tension sensor; 603. End fixing plate; 604. Connector. Detailed Implementation

[0019] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] In the description of this invention, it should be noted that the terms "middle," "upper," "lower," "left," "right," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Example 1

[0021] like Figure 1 As shown, a simulation system for dynamic wake of a wind farm based on digital twins includes a wind farm numerical simulation module, a wind turbine simulation module, and a data visualization module. The wind farm numerical simulation module includes a wind farm inflow turbulence wind speed generation unit and a dynamic model coupling the wind farm wake with the wind turbine transmission chain. The dynamic model is used to output numerical simulation calculation results and issue speed commands. The wind turbine simulation module includes multiple wind turbine load testing platforms. The wind turbine load testing platforms use the speed commands as reference speeds for tracking and output test data. The data visualization module calls the wind farm numerical simulation module and receives the test data. The data visualization module is used to coordinate the wind farm numerical simulation module and the wind turbine simulation module to demonstrate the effect of digital twins.

[0022] This simulation system combines numerical and physical simulations to achieve closed-loop testing of the entire process from turbulent wind farms to power generation. It can quantitatively evaluate the differences in the accuracy of different wake models in predicting downstream unit power, reveal the electromechanical response characteristics of the drive train under dynamic wind speed excitation, and provide an efficient testing solution for wind farm optimization layout, wake control strategy design, and drive train reliability research.

[0023] Existing wind farm wake simulations generally employ numerical calculations, lacking actual data verification from physical experimental platforms. Existing wind turbine drive test benches either lack wake models or only involve static models of the wind farm wake, failing to consider the dynamic characteristics of the wind turbine's drivetrain. This invention employs a digital twin approach, combining the wind farm wake model with a wind turbine (drivetrain six-degree-of-freedom load) test platform to form a closed-loop circuit, enabling a more realistic and accurate simulation of the dynamic wake characteristics of wind turbines.

[0024] This digital twin-based wind farm dynamic wake simulation system can not only generate wind farm wake results calculated by numerical simulation, but also conduct data verification and comparative analysis through the wind turbine load testing platform, taking into account both numerical calculation and measured results, and has good reliability.

[0025] Furthermore, the wind farm inflow turbulent wind speed generation unit can generate the turbulent wind speed inflow into the wind farm based on average wind speed, fluctuating wind speed, and Fourier synthesis method; the wind farm inflow turbulent wind speed generation unit has a variety of turbulent wind speed generation models built in, and different turbulent inflow wind speeds can be obtained by calling the turbulent wind speed generation models according to user needs.

[0026] Furthermore, the dynamic model includes a wind farm wake model calculation unit and a wind turbine drivetrain speed calculation unit. The wind farm wake model calculation unit has multiple wind farm wake calculation models built-in. Different wind turbine wake fields can be obtained by calling these models according to user needs. The wind turbine drivetrain speed calculation unit has wind turbine aerodynamic models, electromechanical control equations, and operating modes built-in. It is used to calculate the drivetrain speeds of multiple wind turbines and obtain the speed commands. The wind turbine drivetrain speed calculation unit can simultaneously calculate the drivetrain speeds of multiple wind turbines. According to user needs, it can call various wind turbine types, capacities, and operating modes (such as optimal power point tracking or variable pitch control models) and generate corresponding drivetrain speeds as commands.

[0027] Furthermore, such as Figure 2 and Figure 3As shown, the wind turbine load test platform includes a wind turbine nacelle 1 under test and a non-contact electromagnetic force loading device 2 connected to the wind turbine nacelle 1. The non-contact electromagnetic force loading device 2 is connected to a drive motor 4 via a permanent magnet eddy current coupler 3. The drive motor 4 is used to load wind torque load and simulate the low-speed operation of the wind turbine. The drive motor 4 adopts variable frequency motor speed control, and uses the speed command issued by the transmission chain speed calculation unit as the reference speed for tracking.

[0028] The non-contact electromagnetic force loading device 2 includes a housing 201 and a flywheel inertia disk 5 suspended within the housing 201. The input shaft of the flywheel inertia disk 5 is connected to the permanent magnet eddy current coupler 3, and the output shaft of the flywheel inertia disk 5 is connected to the wind turbine nacelle 1 under test. A plurality of axially and radially arranged electromagnet assemblies 6 are provided on the outer periphery of the housing 201. Each electromagnet assembly 6 corresponds non-contactly to the axial or radial disk surface of the flywheel inertia disk 5. The combined attraction of the plurality of electromagnet assemblies 6 acts on the flywheel inertia disk as a multi-degree-of-freedom non-torque load.

[0029] The wind turbine load testing platform adopts a six-degree-of-freedom load testing platform solution. This platform not only simulates the flywheel inertia, but also uses electromagnetic force non-contact loading, which has the characteristics of being non-contact and requiring no lubrication, thus eliminating reliability and lifespan issues caused by contact fatigue and wear. At the same time, the flywheel inertia disk is suspended and connected to the drive motor through a permanent magnet eddy current coupler, which can realistically reproduce the actual operating state of the wind turbine and has good accuracy.

[0030] The wind turbine nacelle 1 under test should be a single nacelle unit, including components such as the gearbox, generator, and converter. The flywheel inertia disk 5 is a five-degree-of-freedom non-torque load-bearing component, used to equivalently replace a wind turbine of equal mass and moment of inertia. A certain number of electromagnet assemblies 6 are evenly arranged on both sides and circumferentially of the flywheel inertia disk 5, and the attraction force of each electromagnet assembly 6 is used to synthesize a five-degree-of-freedom non-torque load. When a sufficient number of electromagnet assemblies are set, the requirements for testing the six-degree-of-freedom load of the main drive train of high-power or even ultra-high-power wind turbine units can be met.

[0031] Both the static and dynamic performance of the electromagnetic force loading system are controllable online in real time. When the displacement of the flywheel inertia disk is relatively small, the electromagnetic loading force is linearly proportional to the input control current, exhibiting good linear controllability. Moreover, the displacement and loading force of the flywheel inertia disk can be detected in real time and fed back to the entire simulation system for closed-loop control, thereby further improving the control accuracy of the loading force.

[0032] Furthermore, the electromagnet assembly 6 includes an electromagnet body 601 and a tension sensor 602 connected to the electromagnet body 601. One end of the electromagnet body 601 passes through the housing 201 and corresponds to the disk surface (circumferential or radial) of the flywheel inertia disk 5 with a gap. The tension sensor 602 is located at the end away from the flywheel inertia disk 5 and is connected to an end fixing plate 603. The end fixing plate 603 is connected to the housing 201 by several connectors 604, such as long bolts. This structure allows the electromagnet assembly 6 to be easily installed on the housing 201, while allowing the electromagnet body 601 to pass through the housing 201 without contact and be arranged opposite to the disk surface of the flywheel inertia disk 5.

[0033] The electromagnet assembly 6 uses the attraction force generated between the electromagnet body 601 and the flywheel inertia disk 5 to pull the force sensor 602 to generate a signal, so that the loading force can be detected in real time and fed back to the simulation system for closed-loop control, thereby improving the control accuracy of the loading force. Example 2

[0034] A simulation method for dynamic wake of a wind farm based on digital twins, the simulation method comprising the following steps: Construct the inflow turbulent wind speed generation unit for the wind farm, and call the turbulent wind speed generation model to obtain the turbulent inflow wind speed; A wind farm wake model calculation unit is constructed to calculate the wake wind speed based on the turbulent inflow wind speed. A wind turbine drive train speed calculation unit is constructed, and the dynamic model is obtained by coupling the wind turbine drive train speed calculation unit and the wind farm wake model calculation unit. Based on the dynamic model, the wind turbine main shaft speed and the wake wind speed, the aerodynamic torque of the wind turbine is obtained, the wind turbine drive train speed is calculated, and the speed command is sent to the wind turbine load test platform. The wind turbine load test platform simulates the low-speed operation condition of the wind turbine according to the speed command and outputs the test data. The data visualization module compares and verifies the received test data with the numerical simulation results output by the dynamic model, and corrects the numerical simulation results accordingly. Specifically, the data visualization module can call the wind farm numerical simulation module to generate different dynamic wakes of the wind farm. At the same time, the output parameters such as rotational speed, voltage, and power of the wind turbine simulation module can be fed back to the data visualization module and compared and verified with the calculation results of the wind farm numerical simulation module to correct the numerical simulation results.

[0035] Furthermore, the turbulent inflow wind speed includes the average wind speed and the turbulent wind speed with random characteristics. The turbulent wind speed with random characteristics uses the power spectrum to define the density function and spatial coherence in the frequency domain, and then uses the inverse Fourier transform to generate a time series, so as to obtain the simulated turbulent wind speed of all simulation points in space at each time step.

[0036] A turbulent wind speed generation model is a prerequisite for establishing wind turbine and wind farm models. The inverse Fourier transform method is used to establish the turbulent inflow wind speed model. For example... Figure 4 As shown, this model can generate random, full-field, turbulent inflow wind speeds and uses a frequency domain statistical model to numerically simulate the time series of wind speed vectors at each point in a two-dimensional vertical rectangular grid fixed in space.

[0037] Turbulent inflow wind speed can be divided into two parts: average wind speed and turbulent wind speed with random characteristics. Average wind speed is the portion of wind speed amplitude that remains constant over a period of 10 minutes or more. Average wind speed can be calculated using exponential law (PL), logarithmic law (LOG), IEC, and low-level jet stream (JET) wind profiles. Turbulent wind speed is generated by irregular wind motion and is often assumed to be a stationary random process in engineering, meaning that the probability distribution of wind speed at any given time is exactly the same. Therefore, time-domain simulation of fluctuating wind speed can be performed using power spectral functions. These include power spectral models such as IEC Kaimal (IECKAI), IEC von Karman (IECVKM), Risø Smooth Terrain (SMOOTH), and four site-specific models (NWTCUP, GP_LLJ, WF_UPW, WF_07D, and WF_14D). By using the power spectrum to define the density function and spatial coherence in the frequency domain, and then using the inverse fast Fourier transform to generate a time series, the turbulent wind speeds simulated at all simulated points in space at each time step are obtained.

[0038] Furthermore, based on the turbulent wind speed generation model, an aerodynamic model of the wind turbine is established using blade element momentum theory, and a transmission chain equation of the wind turbine is established using a dual mass block model, coupling the aerodynamic model and the transmission chain equation. On this basis, considering the wake effect of the wind farm and the lag time or propulsion time of the incoming wind arriving at the subsequent wind turbine, a dynamic model based on the dynamic coupling of the wind farm, the aerodynamics of the wind turbine, and the transmission chain is established.

[0039] Specifically, based on the aforementioned turbulent wind speed generation model, the aerodynamic performance model of the wind turbine is established using blade element momentum theory (BEM). BEM theory is a widely used method in wind turbine design and performance analysis. This theory assumes that the blade can be divided into multiple relatively independent blade elements along its span. Therefore, each blade element can be used for two-dimensional aerodynamic performance calculations on the airfoil plane. The inflow angle of the blade element can be calculated from the inflow wind speed, rotational speed, and axial and tangential induced velocities. Based on BEM theory, when the lift coefficient is known...C l drag coefficient C d turbulent inflow wind speed v and rotational speed ω At this point, the aerodynamic parameters of each blade element can be solved iteratively. Before that, corrections for special positions and operating conditions need to be considered, such as tip and hub loss corrections, Glauber's corrections for high induced velocities, and corrections for skew wakes.

[0040] Then, based on the aerodynamic parameters of each blade element, the force and torque of each blade element can be solved, and the aerodynamic torque of the wind turbine can be obtained by integrating it along the spanwise direction of the blade. This torque is then used as the input to the aerodynamic part of the wind turbine's transmission chain model. The transmission chain equations of the wind turbine are established using a dual-mass block model, and a model of the variable pitch mechanism is established. In the low wind speed range below the rated wind speed, the tip speed ratio is stabilized near its optimal value using a proportional-integral (PI) optimal power point tracking (EPR) control method. In the high wind speed range above the rated wind speed, the output power is stabilized near its rated value using a proportional power stability control method.

[0041] To couple the transmission chain equations and the aerodynamic model, the BEM theory can be abstracted into a nonlinear algebraic equation. ,in, T a The aerodynamic torque is represented by the transmission chain model, which is a typical first-order equation. An implicit system of equations is constructed by combining the two equations.

[0042] ......(1) Based on this, considering the wake effect of the wind farm and the lag or propulsion time of the incoming wind reaching the subsequent wind turbines, a dynamic model coupling the aerodynamics of the wind farm and the wind turbine's transmission chain is established. For example... Figure 5 As shown, considering the turbulent inflow wind speed passing sequentially through a series of distributed [wind turbines] along the horizontal axis of the wind turbine. N The coordinates of each typhoon turbine are as follows: x 1, x 2, …, x j , …, x N Take the first j Typhoon generator front-end micro-element segment d x Then there is a infinitesimal time d. t for ......(2) In the formula, coordinates x , No. j The wake wind speed at the typhoon turbine.

[0043] Integrating both sides of equation (2), we can obtain the inflow velocity reaching the first... j The propulsion time of the typhoon turbine is ......(3) The wake velocity can be calculated using an engineering wake model, such as the common Jensen wake model, which is a wake model calculation unit. ......(4) In the formula, The time series of turbulent inflow wind speed, For wind turbines j Previous wind turbines i The set, a i , r i , r w,i Wind turbine i The axial induction factor, rotor radius, and wake spread radius, A j , A ij These are wind turbine units. j wind turbine area and wind turbine i and j The overlapping area of ​​the wind turbines between them.

[0044] The time required for different turbulent inflow wind speeds to reach the wind turbine can be calculated using equations (3) and (4). t 1, t 2,…, t j , …, t N and the corresponding wake velocity value v 1, v 2, …, v j , …, v N When the turbulent inflow wind speed varies with time, let the time series of the turbulent inflow wind speed be... The sampling period is Δ t Then, it is necessary to recalculate the arrival time of each wind turbine according to the new turbulent inflow wind speed, which can be formed in the form of a list, as shown in Table 1.

[0045] Table 1: Calculation of wake propagation time and wake velocity in wind farms

[0046] Table for k The wake of the turbulent inflow wind speed reaches the first j The time of the typhoon generator set.

[0047] Let the current time be T Then we can compare the wake propulsion time with the time in Table 1. T Solve the system of differential-algebraic equations for the dynamic coupling of aerodynamics and transmission chains of each wind turbine, assuming the current time... A set of differential algebraic equations with different integration intervals can be used to describe the dynamics of the wind turbine drive train and the wake propulsion process within a wind farm, i.e., ......(5) In the formula, Wind turbine j The aerodynamic torque, spindle speed and front wake velocity, The wake velocity values ​​can be taken sequentially according to the wake wind speed arrival time in Table 1.

[0048] By solving the dynamic model of equation (5) and the corresponding aerodynamic model simultaneously, the dynamics of the wind turbine drive train and the wake process of the wind farm can be described from a holistic perspective.

[0049] Furthermore, the wake calculation model includes the Howland, Gebraad, or Frandsen wake model, and the output test data includes parameter data such as output speed, voltage, and electric power.

[0050] Furthermore, the issuance of the speed command and the feedback of the test data can be achieved through fieldbus communication, such as using the Modbus RTU communication protocol to encapsulate the command into a hexadecimal command frame, and sending the command to the wind turbine simulation module in a cyclic manner at time step intervals through the serial communication module.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A simulation system for dynamic wake of a wind farm based on digital twins, characterized in that, The system includes a wind farm numerical simulation module, a wind turbine simulation module, and a data visualization module. The wind farm numerical simulation module includes a wind farm inflow turbulence wind speed generation unit and a dynamic model that couples the wind farm wake with the wind turbine transmission chain. The dynamic model is used to output numerical simulation calculation results and issue speed commands. The wind turbine simulation module includes multiple wind turbine load testing platforms. The wind turbine load testing platforms use the speed commands as reference speeds for tracking and output test data. The data visualization module calls the wind farm numerical simulation module and receives the test data. The data visualization module is used to coordinate the wind farm numerical simulation module and the wind turbine simulation module to demonstrate the effect of digital twin.

2. The simulation system for dynamic wake of a wind farm based on digital twins according to claim 1, characterized in that, The wind farm inflow turbulent wind speed generation unit has multiple built-in turbulent wind speed generation models, and different turbulent inflow wind speeds can be obtained by calling the turbulent wind speed generation models.

3. The simulation system for dynamic wake of a wind farm based on digital twins according to claim 1, characterized in that, The dynamic model includes a wind farm wake model calculation unit and a wind turbine drivetrain speed calculation unit. The wind farm wake model calculation unit has multiple wind farm wake calculation models built-in. Different wind turbine wake fields are obtained by calling the wind farm wake calculation models. The wind turbine drivetrain speed calculation unit has wind turbine aerodynamic models, electromechanical control equations and operating modes built-in. It is used to calculate the drivetrain speed of multiple wind turbines and obtain the speed command.

4. The simulation system for dynamic wake of a wind farm based on digital twins according to claim 1, characterized in that, The wind turbine load test platform includes a wind turbine nacelle under test and a non-contact electromagnetic force loading device connected to the wind turbine nacelle under test. The non-contact electromagnetic force loading device is connected to a drive motor through a permanent magnet eddy current coupler. The drive motor is used to load wind torque load and simulate the low-speed operation of the wind turbine.

5. The simulation system for dynamic wake of a wind farm based on digital twins according to claim 4, characterized in that, The non-contact electromagnetic force loading device includes a housing and a flywheel inertia disk suspended within the housing. The input shaft of the flywheel inertia disk is connected to the permanent magnet eddy current coupler, and the output shaft of the flywheel inertia disk is connected to the wind turbine nacelle under test. The outer periphery of the housing is provided with a plurality of axially and radially arranged electromagnet assemblies. Each electromagnet assembly corresponds non-contactly to the axial or radial disk surface of the flywheel inertia disk. The combined attraction of the plurality of electromagnet assemblies acts on the multi-degree-of-freedom non-torque load of the flywheel inertia disk.

6. The simulation method for a simulation system of dynamic wake of a wind farm based on digital twins according to any one of claims 1 to 5, characterized in that, The simulation method includes the following steps: Construct the inflow turbulent wind speed generation unit for the wind farm, and call the turbulent wind speed generation model to obtain the turbulent inflow wind speed; Construct a wind farm wake model calculation unit to calculate the wake wind speed based on the turbulent inflow wind speed; A wind turbine drive train speed calculation unit is constructed, and the dynamic model is obtained by coupling the wind turbine drive train speed calculation unit and the wind farm wake model calculation unit. Based on the dynamic model, the main shaft speed of the wind turbine and the wake wind speed, the aerodynamic torque of the wind turbine is obtained, the drive train speed of the wind turbine is calculated, and the speed command is sent to the wind turbine load test platform. The wind turbine load test platform simulates the low-speed operation condition of the wind turbine according to the speed command and outputs the test data. The data visualization module compares and verifies the received test data with the numerical simulation results output by the dynamic model, and corrects the numerical simulation results accordingly.

7. The simulation method for a wind farm dynamic wake simulation system based on digital twins according to claim 6, characterized in that, The turbulent inflow wind speed includes the average wind speed and the turbulent wind speed with random characteristics. The turbulent wind speed with random characteristics uses the power spectrum to define the density function and spatial coherence in the frequency domain, and then uses the inverse Fourier transform to generate a time series, so as to obtain the simulated turbulent wind speed of all simulation points in space at each time step.

8. The simulation method for a wind farm dynamic wake simulation system based on digital twins according to claim 6, characterized in that, Based on the turbulent wind speed generation model, an aerodynamic model of the wind turbine is established using blade element momentum theory, and the transmission chain equation of the wind turbine is established using a dual-mass block model, coupling the aerodynamic model and the transmission chain equation. On this basis, considering the wake effect of the wind farm and the lag or propulsion time of the incoming wind reaching subsequent wind turbines, a dynamic model is established based on the coupling of the wind farm, wind turbine aerodynamics, and transmission chain dynamics. The following model describes the dynamics of the transmission chain of each wind turbine within the wind farm and the wake propulsion process: ， In the formula, Wind turbine j The aerodynamic torque, spindle speed and front wake velocity, t 1, t 2, …, t j , …, t N T represents the time required for different turbulent inflow wind speeds to reach the wind turbine sequentially, where T is the current time.

9. The simulation method for a wind farm dynamic wake simulation system based on digital twins according to claim 6, characterized in that, The wake calculation model includes the Howland, Gebraad, or Frandsen wake model, and the output test data includes parameter data such as output speed, voltage, and electric power.

10. The simulation method for a wind farm dynamic wake simulation system based on digital twins according to claim 6, characterized in that, The issuance of the rotational speed command and the feedback of the test data are achieved by encapsulating the command into a hexadecimal command frame using a communication protocol, and sending the command to the wind turbine simulation module in a cyclic manner at time step intervals through the serial communication module.

Images