Floating wind turbine digital twin based on integrated simulation and control method and system

Abstract

The application belongs to the technical field of ocean engineering and wind power generation, and discloses a digital twin and control method and system of a floating wind turbine based on integrated simulation, comprising: obtaining real-time environmental parameters, structural response parameters, wind turbine motion state parameters and wind turbine operation state parameters through multiple types of sensors on a physical entity floating wind turbine; pre-processing, feature extraction and fusion processing of the obtained multiple types of sensor data; based on the fusion-processed data, driving a simulation model that is real-time mapped with the physical entity floating wind turbine and is constructed and run in a digital space; based on the simulation results and prediction data of the simulation model, assessing and predicting the resilience of the physical entity floating wind turbine; according to the assessment and prediction results, using a controller to perform closed-loop adjustment on the physical entity floating wind turbine. The application improves the state monitoring and prediction capability in the operation and maintenance process of the floating wind turbine, thereby improving the operation stability of the floating wind turbine.

Description

Technical Field

[0001] This invention relates to the fields of marine engineering and wind power generation technology, and in particular to a digital twin and control method and system for floating wind turbines based on integrated simulation. Background Technology

[0002] With onshore and near-shore wind power resources nearing saturation, deep-sea wind power development has become an inevitable trend. Among these, floating wind turbines, using offshore platform foundations that float on the water surface instead of fixed pile foundations, offer better adaptability to deep-water environments and are an important means of developing and utilizing deep-sea wind energy. Due to the complex marine environment, floating wind turbines are subjected to complex loads such as cyclic aerodynamic loads, hydrodynamic loads, structural elastic forces, mooring forces, and servo control forces. This results in highly complex states for floating wind turbines, high operational safety risks, and difficulties in achieving high-precision integrated simulation and developing efficient prediction and control strategies.

[0003] Existing coupled simulations of floating wind turbines mostly employ traditional rigid-flexible coupling methods, which have a fixed load transfer direction and struggle to fully capture the complex real-time interactions between aerodynamics, hydrodynamics, servo control, and structural elasticity, resulting in accumulated computational errors. Furthermore, traditional rigid-flexible coupling simulation methods suffer from inherent limitations in their computational principles when predicting structural response and motion attitude under extreme conditions, making it difficult to eliminate safety risks and uncertainties in operation and maintenance decisions.

[0004] Furthermore, traditional simulation methods are typically based on offline design models, resulting in significant data lag and an inability to effectively capture transient responses and dynamic coupling processes. This leads to substantial discrepancies between the simulation model and the actual operating conditions. During the operation phase, the lack of precise model-based control strategies prevents the implementation of dynamic optimization control based on real-world conditions, thus limiting the power generation efficiency and operational safety of wind turbines in complex sea conditions.

[0005] Due to the above shortcomings, existing floating wind turbine simulation methods heavily rely on feedback regulation in terms of control strategies, lacking feedforward mechanisms. They often intervene only after significant changes in the turbine's attitude, failing to achieve active stability control based on motion prediction. This leads to insufficient state monitoring and prediction capabilities during the operation and maintenance of floating wind turbines due to the disconnect between the simulation model and the physical entity.

[0006] Therefore, existing technologies still need improvement. Summary of the Invention

[0007] The technical problem to be solved by the present invention is that, in view of the defects of the prior art, the present invention provides a digital twin and control method and system for floating wind turbines based on integrated simulation, so as to solve the problem of insufficient state monitoring and prediction capabilities caused by the disconnect between the simulation model and the physical entity during the operation and maintenance of floating wind turbines.

[0008] The technical solution adopted by this invention to solve the technical problem is as follows: In a first aspect, the present invention provides a digital twin and control method for floating wind turbines based on integrated simulation, comprising: Real-time environmental parameters, structural response parameters, wind turbine motion status parameters, and wind turbine operating status parameters are acquired through various types of sensors on the physical floating wind turbine. Preprocessing, feature extraction, and fusion processing are performed on the acquired multi-type sensor data; Based on the fused data, a high-fidelity integrated simulation model covering aerodynamics, hydrodynamics, servo control, and structural elastic coupling is driven; wherein, the simulation model is a simulation model constructed and run in digital space that is mapped in real time to the physical entity floating wind turbine. Based on the simulation results and prediction data of the simulation model, the toughness of the physical entity floating wind turbine is evaluated and predicted. Based on the assessment and prediction results, the physical entity floating wind turbine is subjected to closed-loop regulation using a controller.

[0009] In one implementation, acquiring real-time environmental parameters, structural response parameters, wind turbine motion state parameters, and wind turbine operating state parameters through multiple types of sensors on the physical floating wind turbine includes: The real-time environmental parameters are obtained by acquiring wind speed and wave data using an anemometer and radar, respectively. The structural response parameters were obtained using fiber optic strain gauges. The motion state parameters of the wind turbine are obtained through the global navigation satellite system receiver and inertial measurement unit at the platform's center of mass and key nodes; The wind turbine operating status parameters are obtained by monitoring the power generation, blade pitch angle, nacelle yaw angle, and generator speed using wind turbine operating status monitoring sensors.

[0010] In one implementation, the preprocessing, feature extraction, and fusion processing of the acquired multi-type sensor data includes: The data from the various types of sensors are time-stamp aligned and cleaned to obtain pre-processed data; Construct a deep learning feature extraction model; wherein, the deep learning feature extraction model is a hybrid model of multi-channel convolutional neural network and long short-term memory network; Based on the deep learning feature extraction model, spatial and temporal features are extracted from the preprocessed data, and after fusion processing, a denoised high-dimensional feature vector is output; wherein, the high-dimensional feature vector is the benchmark data used for simulation model correction.

[0011] In one implementation, the step of driving a high-fidelity integrated simulation model encompassing aerodynamics, hydrodynamics, servo control, and structural elastic coupling based on the fused data includes: The high-fidelity integrated simulation model is generated based on the design parameters of the physical floating wind turbine. Based on the fused data, the simulation model is driven to perform simulation and prediction to obtain the simulation results and prediction data; wherein, the prediction data is the prediction results of the structural response and motion state of the physical entity floating wind turbine within a preset time period.

[0012] In one implementation, the step of driving the simulation model to perform simulation and prediction based on the fused data to obtain the simulation results and prediction data includes: The key parameters in the simulation results are compared with the measured data in the multi-type sensor data, and the key parameters of the simulation model are dynamically adjusted using a parameter identification algorithm to minimize the error between the model output and the actual behavior of the physical entity.

[0013] In one implementation, the assessment and prediction of the resilience of the physical floating wind turbine based on the simulation results and prediction data of the simulation model includes: Obtain the initial stability parameters and the corresponding safe ranges of key parameters of the physical floating wind turbine; The simulation results are compared with the safety range to assess the safety status of the physical floating wind turbine at future moments and generate a resilience assessment report. The dimensions of the resilience assessment include any one or more combinations of tilt margin, recovery efficiency, and preset performance benchmark risk indicators.

[0014] In one implementation, the step of using a controller to perform closed-loop regulation of the physical floating wind turbine based on the evaluation and prediction results includes: Based on the evaluation and prediction results, control commands and corresponding adjustment strategies for the controller are generated; wherein, the adjustment strategies include any one or more combinations of blade collective pitch strategy, independent pitch strategy, ballast control strategy, and shutdown protection strategy; The physical floating wind turbine is subjected to closed-loop regulation according to the control command and the corresponding adjustment strategy to maintain the attitude of the physical floating wind turbine within the preset power generation range.

[0015] Secondly, the present invention provides a digital twin and control system for floating wind turbines based on integrated simulation, comprising: The physical sensing module is used to acquire real-time environmental parameters, structural response parameters, wind turbine motion status parameters, and wind turbine operating status parameters through various types of sensors on the physical floating wind turbine. The data fusion module is used for preprocessing, feature extraction, and fusion processing of acquired multi-type sensor data; The virtual mapping module is used to drive a high-fidelity integrated simulation model that covers aerodynamics, hydrodynamics, servo control, and structural elastic coupling based on the fused data; wherein, the simulation model is a simulation model that is constructed and run in digital space and mapped in real time to the physical entity floating wind turbine. The evaluation and prediction module is used to evaluate and predict the toughness of the physical entity, the floating wind turbine, based on the simulation results and prediction data of the simulation model. The decision control module is used to perform closed-loop regulation of the physical entity, the floating wind turbine, based on the evaluation and prediction results using the controller.

[0016] Thirdly, the present invention provides a terminal comprising: a processor and a memory, the memory storing a floating wind turbine digital twin and control program based on integrated simulation, the floating wind turbine digital twin and control program based on integrated simulation being executed by the processor to implement the operation of the floating wind turbine digital twin and control method based on integrated simulation as described in the first aspect.

[0017] Fourthly, the present invention also provides a computer-readable storage medium storing a floating wind turbine digital twin and control program based on integrated simulation, wherein the floating wind turbine digital twin and control program based on integrated simulation, when executed by a processor, is used to implement the operation of the floating wind turbine digital twin and control method based on integrated simulation as described in the first aspect.

[0018] The present invention, by employing the above technical solution, has the following effects: 1) This invention constructs a complete closed loop of "perception-fusion-mapping-evaluation-control", realizing dynamic, bidirectional real-time interaction and optimization between digital virtual entities and physical entities, solving the problem of disconnection between traditional simulation models and physical entities, and greatly improving the synchronicity and accuracy of state monitoring, prediction and control.

[0019] 2) This invention fully considers the strong coupling effect of multiple physical fields such as aerodynamics, hydrodynamics, servo control and structural elasticity, so that the digital twin can truly reflect the nonlinear dynamic response of the floating wind turbine in the complex marine environment, laying a reliable foundation for accurate prediction and evaluation.

[0020] 3) This invention integrates deep learning and physical models, achieving the complementary advantages of data-driven and mechanism-based models, enabling the twin to have self-learning and self-adaptive capabilities.

[0021] 4) This invention can not only assess the current security status, but also realize the transformation of operation and maintenance strategy from post-event response to pre-event prevention based on future environment prediction and advanced simulation.

[0022] 5) Based on the resilience assessment results, this invention can automatically generate and execute the optimal control strategy, actively controlling the platform movement within the optimal power generation range, thus ensuring the stability of system operation. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0024] Figure 1 This is a flowchart of the digital twin and control method for floating wind turbines based on integrated simulation in this invention.

[0025] Figure 2 This is a digital twin diagram of the physical entity floating wind turbine in this invention.

[0026] Figure 3 This is an overall framework diagram of the floating wind turbine digital twin system based on integrated simulation in this invention.

[0027] Figure 4 This is a data flow diagram of the floating wind turbine digital twin system based on integrated simulation in this invention.

[0028] Figure 5 This is a functional schematic diagram of the terminal in one implementation of the present invention.

[0029] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, 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 of the invention and are not intended to limit the invention.

[0031] Exemplary methods Existing floating wind turbine simulation methods heavily rely on feedback regulation in terms of control strategies, lacking feedforward mechanisms. They often intervene only after significant changes in the turbine's attitude occur, failing to achieve active stability control based on motion prediction. This leads to a disconnect between the simulation model and the physical entity during the operation and maintenance of floating wind turbines, resulting in insufficient state monitoring and prediction capabilities.

[0032] To address the above-mentioned technical problems, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation. The method includes: acquiring real-time environmental parameters, structural response parameters, turbine motion state parameters, and turbine operating state parameters using multiple types of sensors on the physical floating wind turbine; preprocessing, extracting features, and fusing the acquired sensor data; driving a simulation model that is constructed and runs in digital space and is mapped in real-time to the physical floating wind turbine based on the fused data; evaluating and predicting the resilience of the physical floating wind turbine based on the simulation results and prediction data of the simulation model; and using a controller to perform closed-loop regulation of the physical floating wind turbine according to the evaluation and prediction results. This invention improves the state monitoring and prediction capabilities during the operation and maintenance of floating wind turbines, thereby improving the operational stability of the floating wind turbines.

[0033] like Figure 1 As shown in the figure, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation, including the following steps: Step S100: Real-time environmental parameters, structural response parameters, wind turbine motion state parameters, and wind turbine operating state parameters are acquired through various types of sensors on the physical floating wind turbine.

[0034] In this embodiment, to implement the above method, a digital twin and control system for floating wind turbines based on integrated simulation is also provided. This system collects real-time operating status data of the floating wind turbine through a multi-source sensor network, utilizes deep learning technology to fuse and extract features from the multi-source heterogeneous data, and constructs a simulation model that is synchronized with the physical entity in real time based on a high-fidelity integrated numerical simulation engine. Through model calibration and advanced prediction, a quantitative assessment of the "toughness index" of the floating wind turbine is achieved, and active adjustments are made based on the assessment results through control strategies such as pitch control and ballast adjustment, thereby improving the stability, safety, and power generation efficiency of the wind turbine under harsh sea conditions.

[0035] In this embodiment, as Figure 3 As shown, the floating wind turbine digital twin and control system achieves, through the construction of a real-time mapping mechanism between physical entities and simulation models, the following: Figure 2 The digital twin of the physical entity floating wind turbine shown.

[0036] As an example, in this embodiment, the floating wind turbine digital twin and control system comprises five main modules: a physical sensing layer, a data fusion layer, a virtual mapping layer, an evaluation and prediction layer, and a decision control layer. (1) Physical sensing layer: Multiple types of sensor networks are deployed on the physical floating wind turbine system to obtain real-time environmental parameters, structural response parameters, wind turbine motion state parameters and wind turbine operating state parameters.

[0037] (2) Data fusion layer: The edge computing terminal or cloud platform is used to process the collected multi-type sensor data. The processing includes data preprocessing and deep learning feature extraction.

[0038] (3) Virtual mapping layer: A high-fidelity simulation model of the wind turbine is established using integrated floating wind turbine simulation software. This simulation model covers aerodynamics, hydrodynamics, control servo, and structural elastic coupling.

[0039] (4) Evaluation and prediction layer: Using a deep learning feature extraction model, the environmental parameters for a certain period of time in the future are predicted, the state simulation of the floating wind turbine for a certain period of time in the future is completed, and the resilience assessment and prediction of the physical floating wind turbine is carried out through the simulation results.

[0040] (5) Decision control layer: Based on the resilience assessment results, control commands are sent to the physical floating wind turbine to achieve closed-loop regulation.

[0041] In this embodiment, based on the floating wind turbine digital twin and control system, real-time environmental parameters, structural response parameters, wind turbine motion state parameters, and wind turbine operating state parameters are first obtained through the physical sensing layer. These data can be used for subsequent data fusion, thereby serving as the benchmark data for the simulation model.

[0042] Specifically, in one implementation of this embodiment, step S100 includes the following steps: Step S101: Obtain wind speed and wave data using an anemometer and radar respectively to obtain the real-time environmental parameters; Step S102: Obtain the structural response parameters using fiber optic strain gauges; Step S103: Obtain the motion state parameters of the wind turbine through the global navigation satellite system receiver and inertial measurement unit of the platform's center of mass and key nodes; Step S104: The wind turbine operating status parameters are obtained by monitoring the power generation, blade pitch angle, nacelle yaw angle and generator speed through the wind turbine operating status monitoring sensor.

[0043] In this embodiment, the multi-type sensor network in the physical sensing layer includes: The environmental parameter monitoring unit includes an anemometer installed in the cabin and a radar for monitoring waves; The structural response parameter monitoring unit includes fiber optic strain gauges arranged at the base of the tower and the base of the blades. The wind turbine motion status parameter monitoring unit includes a GNSS / RTK (a high-precision positioning technology based on the global navigation satellite system) receiver and an inertial measurement unit (IMU) installed at the platform's center of mass and key nodes; The wind turbine operating status parameter monitoring unit is used to monitor power generation, blade pitch angle, nacelle yaw angle, and generator speed.

[0044] Specifically, during environmental parameter monitoring, an anemometer is used to monitor wind speed and direction, and wave radar is used to monitor wave height and wave period to obtain real-time environmental parameters. During structural response parameter monitoring, fiber optic strain gauges are deployed at the base of the tower and blades to obtain real-time stress data of the structure, thus obtaining structural response parameters. During wind turbine motion state parameter monitoring, a global navigation satellite system and inertial measurement unit installed at the platform's center of mass and key nodes are used to obtain the displacement, velocity, and acceleration of the platform's six degrees of freedom motion, thus obtaining wind turbine motion state parameters. During wind turbine operating state parameter monitoring, parameters such as power generation, blade pitch angle, nacelle yaw angle, and generator speed are monitored to obtain wind turbine operating state parameters.

[0045] like Figure 1 As shown in the figure, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation, including the following steps: Step S200 involves preprocessing, feature extraction, and fusion processing of the acquired multi-type sensor data.

[0046] In this embodiment, after acquiring multi-type sensor data based on the physical sensing layer, these multi-type sensor data are deeply fused through a data fusion layer. Specifically, the data fusion layer is used for: The sensor data with different sampling frequencies are timestamped and cleaned. Based on the preprocessed data, the key parameters of the integrated simulation model in the virtual mapping layer are dynamically adjusted through a parameter identification algorithm to minimize the error between the model output and the real behavior of the physical entity.

[0047] The data fusion layer is also used to construct a deep learning feature extraction model, which is a hybrid model of multi-channel convolutional neural network and long short-term memory network. This model is used to extract spatial and temporal features from multi-source heterogeneous raw sensor data and output a noise-reduced high-dimensional feature vector as the benchmark data for model correction in the virtual mapping layer.

[0048] Specifically, in one implementation of this embodiment, step S200 includes the following steps: Step S201: Perform timestamp alignment and cleaning on the multi-type sensor data to obtain preprocessed data; Step S202: Construct a deep learning feature extraction model; wherein the deep learning feature extraction model is a hybrid model of a multi-channel convolutional neural network and a long short-term memory network; Step S203: Based on the deep learning feature extraction model, spatial and temporal features are extracted from the preprocessed data, and after fusion processing, a denoised high-dimensional feature vector is output; wherein, the high-dimensional feature vector is the benchmark data used for simulation model correction.

[0049] In this embodiment, the data fusion method can be further subdivided into the following methods: 1. After receiving data from multiple types of sensors, perform data preprocessing, including timestamp alignment and cleaning of data from different sampling frequencies.

[0050] 2. By using a parameter identification algorithm, key parameters in the simulation model are automatically adjusted to minimize the error between the simulation model's output and the actual behavior of the physical entity.

[0051] 3. Construct a hybrid model of multi-channel convolutional neural network (CNN) and long short-term memory network (LSTM), input multi-source heterogeneous raw data, use CNN to extract spatial features, use LSTM to extract time series features, and finally output a denoised high-dimensional feature vector as the benchmark data for simulation model correction.

[0052] Specifically, the steps for adjusting key parameters in the simulation model include: 1) Use AKF (Enhanced Kalman Filter) to adjust key parameters in the simulation model to minimize the error between the simulation model's output and the actual behavior of the physical entity. The adjustment method is as follows: State prediction: ; Observation and prediction: ; New interest calculation: ; Status Update: ; in, This represents the k-th predicted value for each parameter; This represents the k-th monitored value for each parameter; This represents the state transition matrix for each parameter; A mapping operator that represents the relationship between the sensor output signal and the physical state of the parameters; The Kalman gain matrix is ​​represented by... and get.

[0053] 2) The parameters to be adjusted include the following types: environmental parameters, fan motion parameters, and structural response parameters.

[0054] The environmental parameters include two parts: aerodynamic calculation parameters above the water surface and fluid dynamic calculation parameters below the water surface. The aerodynamic calculation parameters include wind speed. The fluid dynamic calculation parameters include wave height, wave period and flow direction, and wave flow direction parameters. Among them, the wave flow direction parameter uses the angle between the current flow direction and the initial flow direction.

[0055] The fan motion parameters are expressed as follows: ; Among them, x, y, z, θ and ψ represent swaying, rolling, heaving, swaying, and pitching.

[0056] The structural response parameters include: heave acceleration, tension of each mooring line, and tower base bending moment. The tower base bending moment is indirectly calculated based on the strain at the tower base, with the adjustment parameter being the strain at the four directions of the tower base.

[0057] 3) For the above multi-source heterogeneous data, a multi-channel parallel input matrix is ​​constructed, with each channel corresponding to a type of sensor data. The data is then standardized and normalized to eliminate the influence of different physical quantities on the calibration accuracy through a normalization operator.

[0058] The processed standardized data enters the spatial feature extraction layer. A one-dimensional convolutional layer is used to scan features along the time step. Multiple sets of convolutional kernels are used to extract spatial coupling features of different frequency bands, capturing the physical correlation between platform motion and tower stress distribution, and transforming discrete sensor signals into spatial feature vectors that characterize the overall attitude of the wind turbine.

[0059] The output of the CNN is used as the sequence input to the LSTM. Its gating mechanism is used to process the data to extract the system inertial trend from the wind speed evolution history, nonlinear wave load, and structural response characteristics, while filtering out high-frequency random noise and retaining long-term dependency information.

[0060] like Figure 1 As shown in the figure, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation, including the following steps: Step S300: Based on the fused data, drive a high-fidelity integrated simulation model that encompasses aerodynamics, hydrodynamics, servo control, and structural elastic coupling; wherein, the simulation model is a simulation model constructed and run in digital space that is mapped in real time to the physical entity, the floating wind turbine.

[0061] In this embodiment, after fusing data from multiple types of sensors through a data fusion layer, a virtual mapping layer is used to perform real-time mapping between the physical floating wind turbine and the simulation model.

[0062] The virtual mapping layer is specifically used for: generating an integrated simulation model based on the design parameters of the physical entity, the floating wind turbine; receiving real-time data processed by the data fusion layer, driving the integrated simulation model to perform simulation, and outputting the predicted results of the structural response and motion state of the floating wind turbine within a preset time (e.g., 600s); comparing the key parameters output by the simulation with the measured data from the sensors, and using the optimization algorithm of the data fusion layer to adjust the key parameters of the integrated simulation model in reverse, thereby achieving real-time synchronous mapping between the simulation model and the physical entity.

[0063] The integrated simulation model in the virtual mapping layer is a numerical model built on the OpenFAST (open source wind turbine simulation tool) platform or the SAToe (offshore floating wind turbine multi-field coupling numerical tool) platform, which couples the aerodynamic module, hydrodynamic module, servo control module, structural elasticity module and mooring system module.

[0064] Specifically, in one implementation of this embodiment, step S300 includes the following steps: Step S301: Generate the high-fidelity integrated simulation model based on the design parameters of the physical floating wind turbine; Step S302: Based on the fused data, drive the simulation model to perform simulation and prediction to obtain the simulation results and prediction data; wherein, the prediction data is the prediction results of the structural response and motion state of the physical entity floating wind turbine within a preset time period in the future; Step S303: Compare the key parameters in the simulation results with the measured data in the multi-type sensor data, and use the parameter identification algorithm to dynamically adjust the key parameters of the simulation model in order to minimize the error between the model output and the actual behavior of the physical entity.

[0065] In this embodiment, during the generation of a high-fidelity integrated simulation model, the design parameters of the physical floating wind turbine are first obtained. This embodiment uses a semi-submersible floating wind turbine as an example; these parameters are shown in the table below:

[0066] In this embodiment, the floating wind turbine is defined as a multi-rigid-body structure, and an integrated rigid-body simulation model of the floating wind turbine is generated in an open-source aerodynamic-hydraulic-servo-elastic coupling tool (such as openFAST).

[0067] To achieve real-time performance in digital twins, simplified simulations using offline computing in conjunction with online models are employed. Specifically, the workflow of the offline computing module is as follows: S1: Environment Envelope Domain Construction: Acquire site meteorological data and historical unit operating data, and extract spatial environmental characteristic parameters of the target site; A safety margin factor, pre-selected as 1.5, is introduced to dynamically adjust the upper and lower limits of the interval, forming a closed-loop simulation environment parameter constraint envelope domain.

[0068] S2: Offline characterization of dynamic features: Sample points are sampled within the envelope domain, and hydrodynamic characteristic coefficients under different combinations of wind speed, wave period, and wave direction are pre-calculated in the aerodynamic-hydraulic-servo-elastic coupling tool. These coefficients include the added mass matrix, radiation damping matrix, and wave excitation force transfer function. The calculation results are then structured and stored as a multidimensional mapping lookup table.

[0069] S3: Structural dynamics order reduction: During offline computation, a full-field snapshot of the system's structural response is synchronously acquired. The principal mode basis vector space of the system is extracted using the intrinsic orthogonal decomposition (POD) technique, thereby transforming the high-dimensional dynamic equations into low-dimensional modal evolution equations and constructing a reduced-order computational model (ROM) for the real-time engine.

[0070] The simulation process for a certain duration of the online computation model is as follows: S1: Real-time environmental constraints: The current external environmental parameter vector is obtained in real time by a sensor array deployed on the floating wind turbine.

[0071] By using a multidimensional linear interpolation algorithm, a fast index is performed in the offline stored multidimensional mapping lookup table to instantly obtain the hydrodynamic load coefficient and aerodynamic load coefficient under the current constraints.

[0072] S2: Dynamic evolution within the reduced-order space: Based on the POD method, the acquired real-time load coefficients are projected onto an offline-generated low-dimensional basis vector space, simplifying the equations into a low-dimensional ordinary differential equation system about the modal coefficients. The simplification method is as follows: ; in, This represents the modal coefficient vector, and its dimension is... ( (representing the original degrees of freedom). , , Represents the reduced-order mass, damping, and stiffness matrices; Time-domain integration is performed using a certain step size (preset to 0.01s).

[0073] S3: Time schedule calculation for preset time length: Using the current sensor feedback state as the initial value, continuous rolling calculations are performed within one calculation cycle (preset to 600s); During the calculation process, the platform displacement output by the simulation is compared with the actual displacement measured by the sensor at fixed intervals. The initial state of the next stage is calibrated in real time using the error correction factor to prevent numerical drift caused by long-endurance calculation and ensure the stability of the simulation under long-endurance conditions.

[0074] S4: Linearly reconstruct the calculated modal coefficients and offline basis vectors to restore the displacement, stress and load distribution of the physical entity in the entire field; transmit the calculation results to the visualization platform in real time to realize the synchronous operation and state prediction of the digital twin and the physical entity.

[0075] like Figure 1 As shown in the figure, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation, including the following steps: Step S400: Based on the simulation results and prediction data of the simulation model, the toughness of the physical entity floating wind turbine is evaluated and predicted.

[0076] In this embodiment, key parameters of the floating wind turbine's motion state are output to the evaluation, prediction, and decision control layer, enabling an effective comparison between the output of the floating wind turbine simulation model and the actual measured data from the sensors.

[0077] The evaluation and prediction layer is specifically used for: pre-storing the initial stability parameters and safety ranges of key parameters of the floating wind turbine; comparing the simulation results output by the virtual mapping layer with the safety ranges to evaluate the safety status of the floating wind turbine at future moments and generate a resilience evaluation report. The resilience assessment dimensions include at least one of tilt margin, recovery efficiency, and RPN (Risk Priority Number) risk indicators based on PBS (Product Breakdown Structure).

[0078] The assessment and prediction layer is also connected to a deep learning-based prediction module, which is used to predict future environmental parameters based on historical and real-time data, and input the predicted environmental parameters into the virtual mapping layer for advanced simulation, thereby realizing resilience assessment based on future states.

[0079] Specifically, in one implementation of this embodiment, step S400 includes the following steps: Step S401: Obtain the initial stability parameters and the safety range corresponding to the key parameters of the physical floating wind turbine; Step S402: Compare the simulation results with the safety range to assess the safety status of the physical floating wind turbine at future moments and generate a resilience assessment report; wherein, the dimensions of the resilience assessment include any one or more combinations of tilt margin, recovery efficiency and preset performance benchmark risk indicators.

[0080] In this embodiment, the initial stable coordinates and motion parameters of the pre-stored floating wind turbine in the evaluation prediction layer are assessed, and a safety range is set for each key parameter.

[0081] The simulation results are compared with the preset safety range to assess the safety status of the floating wind turbine at future moments and generate an assessment report.

[0082] like Figure 1 As shown in the figure, this invention provides a digital twin and control method for floating wind turbines based on integrated simulation, including the following steps: Step S500: Based on the evaluation and prediction results, the controller is used to perform closed-loop regulation of the physical floating wind turbine.

[0083] In this embodiment, the adjustment strategy corresponding to the control command generated by the decision control layer includes at least one of the following: blade collective pitch strategy, independent pitch strategy, ballast control strategy, and shutdown protection strategy, in order to maintain the wind turbine attitude in the optimal power generation range.

[0084] Specifically, in one implementation of this embodiment, step S500 includes the following steps: Step S501: Generate control commands and corresponding adjustment strategies for the controller based on the evaluation and prediction results; wherein, the adjustment strategies include any one or more combinations of blade collective pitch strategy, independent pitch strategy, ballast control strategy, and shutdown protection strategy. Step S502: Perform closed-loop regulation on the physical floating wind turbine according to the control command and the corresponding adjustment strategy to maintain the attitude of the physical floating wind turbine within the preset power generation range.

[0085] In this embodiment, control commands are sent to the floating wind turbine control system based on the resilience assessment results to achieve closed-loop regulation. The regulation strategies include collective or independent blade pitch control, ballast control, and shutdown protection. Through timely adjustments, the wind turbine's attitude is maintained within the optimal power generation range, achieving stable power generation over a long period.

[0086] As an example, the data flow of the floating wind turbine digital twin system based on integrated simulation in this embodiment is as follows: Figure 4 As shown, taking a semi-submersible floating wind turbine as an example, the uplink data acquisition is achieved by building a multi-source sensor network located on the physical entity, including anemometers, IMUs, tension meters, etc., to collect environmental parameters and floating wind turbine motion state parameters in real time. The data is transmitted to the SCADA (Supervisory Control and Data Acquisition) system inside the digital twin platform through a data transmission method mainly based on fiber optic transmission. The multi-dimensional sensor data is integrated, and the data is preprocessed through a data cleaning method mainly based on enhanced Kalman filtering.

[0087] The preprocessed environmental parameters and floating wind turbine motion state parameters are transmitted to the numerical simulation software, such as openFAST, built into the digital twin platform to perform short-term integrated simulation of the floating wind turbine with a step size of 1s-1min, and output the simulation results of the floating wind turbine motion state.

[0088] Within the aerodynamic-hydraulic-servo-elastic coupling tool, an integrated simulation of a semi-submersible floating wind turbine is performed. The aerodynamic load is solved based on blade element momentum theory (BEM) and the generalized dynamic wake (GDW) principle; the hydrodynamic response is calculated using potential flow theory (PF) and the Morrison equation (ME); the structural dynamics are characterized by rigid body dynamics (R), multibody system (MB), and finite element modal superposition (FEM); and the mooring system is analyzed using the quasi-static (QS) analytical method.

[0089] With an existing database of similar floating wind turbines, an external database is used to train a deep learning model for semi-submersible floating wind turbines by analyzing time series data with LSTM and spatial series data with CNN.

[0090] In addition, the preprocessed environmental parameters and floating wind turbine motion state parameters are transmitted to a medium-deep learning model or a brand-new deep learning model (i.e., a model optimized based on a deep learning model), and trained in stages to form parameter prediction models for 1 hour, 1 day, 1 month, and 1 year. After the prediction results are stable and the fluctuation values ​​are within the error range, the prediction model is started to output the environmental parameters for the future prediction duration.

[0091] The deep learning prediction parameters are input into the integrated simulation software, which outputs the floating wind turbine motion state parameters for predicting the future required prediction time.

[0092] The resilience of floating wind turbines is assessed by predicting their motion state parameters, and an operation strategy feedback report is output. The regulation feedback is then executed by the controller ROSCO (an open-source control tool for wind power generation).

[0093] The resilience assessment is based on the simulated future motion posture, fatigue limit, and mooring safety margin, and the system performs a multi-dimensional evaluation. The evaluation dimensions include inclination margin, recovery efficiency, and the PBS-based RPN risk index.

[0094] After each calculation, the sensor data and the calculation results are compared. The deep learning model is trained using the sensor data as a benchmark to improve the prediction accuracy of the deep prediction model. This step eliminates the deviation between the physical entity and the design model through "virtual-real synchronization" correction.

[0095] This embodiment achieves the following technical effects through the above technical solution: 1) This embodiment constructs a complete closed loop of "perception-fusion-mapping-evaluation-control", realizing dynamic, bidirectional real-time interaction and optimization between digital virtual entities and physical entities, solving the problem of disconnection between traditional simulation models and physical entities, and greatly improving the synchronicity and accuracy of state monitoring, prediction and control.

[0096] 2) This embodiment uses an "integrated simulation" high-fidelity model as the core of the virtual mapping, which fully considers the strong coupling effect of multiple physical fields such as aerodynamics, hydrodynamics, servo control and structural elasticity, so that the digital twin can truly reflect the nonlinear dynamic response of the floating wind turbine in the complex marine environment, laying a reliable foundation for accurate prediction and evaluation.

[0097] 3) This embodiment integrates deep learning and physical models. By using a multi-channel CNN-LSTM hybrid model to intelligently extract features and reduce noise from multi-source heterogeneous data, it not only improves data quality but also uses its output to reverse correct and update physical model parameters, achieving the complementary advantages of data-driven and mechanism-based models, enabling the twin to have self-learning and adaptive capabilities.

[0098] 4) This embodiment achieves a leap from "condition monitoring" to "resilience assessment". The system can not only assess the current safety status, but also quantitatively assess and predict the "resilience" of the wind turbine system in resisting disturbances, maintaining functions, and recovering from failures based on future environment prediction and advanced simulation, realizing the transformation of operation and maintenance strategy from post-event response to pre-event prevention.

[0099] 5) This embodiment forms an intelligent decision-making closed loop oriented towards operation and maintenance efficiency. Based on the resilience assessment results, the system can automatically generate and execute the optimal control strategy, actively controlling the platform's movement within the optimal power generation range, effectively extending the power generation window period, and reducing extreme loads. This maximizes operation and maintenance economy while ensuring safety, providing a technical framework for the long-term reliable and efficient operation of floating wind turbines.

[0100] Exemplary device Based on the above embodiments, the present invention also provides a digital twin and control system for floating wind turbines based on integrated simulation, comprising: The physical sensing module is used to acquire real-time environmental parameters, structural response parameters, wind turbine motion status parameters, and wind turbine operating status parameters through various types of sensors on the physical floating wind turbine. The data fusion module is used for preprocessing, feature extraction, and fusion processing of acquired multi-type sensor data; The virtual mapping module is used to drive a high-fidelity integrated simulation model that covers aerodynamics, hydrodynamics, servo control, and structural elastic coupling based on the fused data; wherein, the simulation model is a simulation model that is constructed and run in digital space and mapped in real time to the physical entity floating wind turbine. The evaluation and prediction module is used to evaluate and predict the toughness of the physical entity, the floating wind turbine, based on the simulation results and prediction data of the simulation model. The decision control module is used to perform closed-loop regulation of the physical entity, the floating wind turbine, based on the evaluation and prediction results using the controller.

[0101] Based on the above embodiments, the present invention also provides a terminal, the principle block diagram of which can be as follows: Figure 5 As shown.

[0102] The terminal includes: a processor, a memory, an interface, a display screen, and a communication module connected via a system bus; wherein, the processor of the terminal provides computing and control capabilities; the memory of the terminal includes a computer-readable storage medium and internal memory; the computer-readable storage medium stores an operating system and computer programs; the internal memory provides an environment for the operation of the operating system and computer programs in the computer-readable storage medium; the interface is used to connect to external devices; the display screen is used to display relevant information; and the communication module is used to communicate with a cloud server or other devices.

[0103] When executed by a processor, this computer program is used to implement the operation of a digital twin and control method for floating wind turbines based on integrated simulation.

[0104] It will be understood by those skilled in the art that Figure 5The schematic diagram shown is merely a partial structural diagram related to the present invention and does not constitute a limitation on the terminal to which the present invention is applied. A specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0105] In one embodiment, a terminal is provided, comprising: a processor and a memory, the memory storing a floating wind turbine digital twin and control program based on integrated simulation, the floating wind turbine digital twin and control program being executed by the processor to implement the operation of the above-described floating wind turbine digital twin and control method based on integrated simulation.

[0106] In one embodiment, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a floating wind turbine digital twin and control program based on integrated simulation, which, when executed by a processor, is used to implement the operation of the above-described floating wind turbine digital twin and control method based on integrated simulation.

[0107] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, database, or other media used in the embodiments provided by this invention can include both non-volatile and volatile memory.

[0108] In summary, this invention provides a digital twin and control method and system for floating wind turbines based on integrated simulation, comprising: acquiring real-time environmental parameters, structural response parameters, turbine motion state parameters, and turbine operating state parameters through multiple types of sensors on the physical floating wind turbine; preprocessing, extracting features, and fusing the acquired multi-type sensor data; driving a simulation model that is constructed and run in digital space and mapped in real time to the physical floating wind turbine based on the fused data; evaluating and predicting the resilience of the physical floating wind turbine based on the simulation results and prediction data of the simulation model; and using a controller to perform closed-loop regulation of the physical floating wind turbine according to the evaluation and prediction results. This invention improves the state monitoring and prediction capabilities during the operation and maintenance of floating wind turbines, thereby improving the operational stability of floating wind turbines.

[0109] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A digital twin and control method for floating wind turbines based on integrated simulation, characterized in that, include: Real-time environmental parameters, structural response parameters, wind turbine motion status parameters, and wind turbine operating status parameters are acquired through various types of sensors on the physical floating wind turbine. The acquired multi-type sensor data undergoes preprocessing, feature extraction, and fusion processing, including: timestamp alignment and cleaning of the multi-type sensor data to obtain preprocessed data; constructing a deep learning feature extraction model; wherein the deep learning feature extraction model is a hybrid model of a multi-channel convolutional neural network and a long short-term memory network; extracting spatial and temporal features from the preprocessed data based on the deep learning feature extraction model, and outputting a denoised high-dimensional feature vector after fusion processing; wherein the high-dimensional feature vector serves as benchmark data for simulation model calibration. Based on the fused data, a high-fidelity integrated simulation model covering aerodynamics, hydrodynamics, servo control, and structural elastic coupling is driven; wherein, the simulation model is a simulation model constructed and run in digital space that is mapped in real time to the physical entity floating wind turbine. Based on the simulation results and prediction data of the simulation model, the toughness of the physical floating wind turbine is evaluated and predicted, including: obtaining the initial stability parameters and the corresponding safe ranges of the key parameters of the physical floating wind turbine; comparing the simulation results with the safe ranges to evaluate the safety status of the physical floating wind turbine at future moments, and generating a toughness assessment report; wherein, the dimensions of the toughness assessment include any one or more combinations of tilt margin, recovery efficiency, and preset performance benchmark risk indicators; Based on the assessment and prediction results, the controller performs closed-loop regulation of the physical floating wind turbine, including: generating control commands and corresponding regulation strategies for the controller based on the assessment and prediction results; wherein, the regulation strategies include any one or more combinations of blade collective pitch strategy, independent pitch strategy, ballast control strategy, and shutdown protection strategy; and performing closed-loop regulation of the physical floating wind turbine based on the control commands and corresponding regulation strategies to maintain the attitude of the physical floating wind turbine within a preset power generation range.

2. The floating wind turbine digital twin and control method based on integrated simulation according to claim 1, characterized in that, The method involves acquiring real-time environmental parameters, structural response parameters, wind turbine motion state parameters, and wind turbine operating state parameters through multiple types of sensors on the physical floating wind turbine, including: The real-time environmental parameters are obtained by acquiring wind speed and wave data using an anemometer and radar, respectively. The structural response parameters were obtained using fiber optic strain gauges. The motion state parameters of the wind turbine are obtained through the global navigation satellite system receiver and inertial measurement unit at the platform's center of mass and key nodes; The wind turbine operating status parameters are obtained by monitoring the power generation, blade pitch angle, nacelle yaw angle, and generator speed using wind turbine operating status monitoring sensors.

3. The floating wind turbine digital twin and control method based on integrated simulation according to claim 1, characterized in that, The aforementioned data, after fusion processing, drives a high-fidelity integrated simulation model encompassing aerodynamics, hydrodynamics, servo control, and structural elastic coupling, including: The high-fidelity integrated simulation model is generated based on the design parameters of the physical floating wind turbine. Based on the fused data, the simulation model is driven to perform simulation and prediction to obtain the simulation results and prediction data; wherein, the prediction data is the prediction results of the structural response and motion state of the physical entity floating wind turbine within a preset time period.

4. The floating wind turbine digital twin and control method based on integrated simulation according to claim 3, characterized in that, The process involves driving the simulation model to perform simulation and prediction based on the fused data, obtaining the simulation results and prediction data, followed by: The key parameters in the simulation results are compared with the measured data in the multi-type sensor data, and the key parameters of the simulation model are dynamically adjusted using a parameter identification algorithm to minimize the error between the model output and the actual behavior of the physical entity.

5. A floating wind turbine digital twin and control system based on integrated simulation, used to implement the floating wind turbine digital twin and control method based on integrated simulation as described in any one of claims 1-4, characterized in that, include: The physical sensing module is used to acquire real-time environmental parameters, structural response parameters, wind turbine motion status parameters, and wind turbine operating status parameters through various types of sensors on the physical floating wind turbine. The data fusion module is used for preprocessing, feature extraction, and fusion processing of acquired multi-type sensor data; The virtual mapping module is used to drive a high-fidelity integrated simulation model that covers aerodynamics, hydrodynamics, servo control, and structural elastic coupling based on the fused data; wherein, the simulation model is a simulation model that is constructed and run in digital space and mapped in real time to the physical entity floating wind turbine. The evaluation and prediction module is used to evaluate and predict the toughness of the physical entity, the floating wind turbine, based on the simulation results and prediction data of the simulation model. The decision control module is used to perform closed-loop regulation of the physical entity, the floating wind turbine, based on the evaluation and prediction results using the controller.

6. A terminal, characterized in that, include: The processor and memory, wherein the memory stores a floating wind turbine digital twin and control program based on integrated simulation, the floating wind turbine digital twin and control program based on integrated simulation being executed by the processor to implement the operation of the floating wind turbine digital twin and control method based on integrated simulation as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a floating wind turbine digital twin and control program based on integrated simulation, which, when executed by a processor, is used to implement the operation of the floating wind turbine digital twin and control method based on integrated simulation as described in any one of claims 1-4.

Images