Offshore floating wind farm yaw cooperative control method and system

Abstract

The application discloses a wind turbine yawing cooperative control method and system of offshore floating wind farm, relates to the technical field of offshore wind power, and comprises the following steps: constructing a digital twin of the offshore floating wind farm to obtain a twin wind farm; calling historical wind condition data locally, solving yawing cooperative strategy in the twin wind farm based on the historical wind condition data, and constructing a benchmark cooperative control library; accessing a remote meteorological forecast center to obtain predicted wind condition data falling into a prediction time window; traversing the benchmark cooperative control library based on the predicted wind condition data, and outputting a benchmark cooperative yawing angle; taking the benchmark cooperative yawing angle as an initial point, performing yawing cooperative control optimization in the twin wind farm according to the predicted wind condition data, and outputting a predicted cooperative yawing angle; and executing power generation intervention on the offshore floating wind farm according to the predicted cooperative yawing angle in the prediction time window. Thus, the technical effects of prospective power generation intervention, improved power generation efficiency and stability are achieved.

Description

Technical Field

[0001] This invention relates to the field of offshore wind power technology, and in particular to a method and system for coordinated control of wind turbine yaw in offshore floating wind farms. Background Technology

[0002] In the operation of offshore floating wind farms, turbine yaw control is crucial for improving power generation efficiency. Traditional yaw control methods for offshore floating wind farms mainly rely on real-time wind condition data to adjust the yaw of individual turbines.

[0003] Traditional offshore floating wind farm turbine yaw control has the following technical drawbacks: First, it relies solely on real-time wind data and cannot fully utilize the patterns and information in historical wind data; second, it cannot effectively combine meteorological forecast data for forward-looking control, and cannot predict in advance the impact of wind changes on the coordinated operation of the entire wind farm turbines, thereby affecting the overall efficiency and equipment safety of the wind farm. Summary of the Invention

[0004] This invention provides a method and system for coordinated yaw control of wind turbines in offshore floating wind farms, in order to solve the technical problems of existing technologies, such as lack of perspective on wind condition changes and historical patterns, single control dimension, and insufficient predictability, and to achieve the technical effects of forward-looking power generation intervention and improved power generation efficiency and stability.

[0005] In a first aspect, the present invention provides a method for coordinated yaw control of wind turbines in an offshore floating wind farm, wherein the method includes:

[0006] A digital twin of an offshore floating wind farm is constructed to obtain a twin wind farm.

[0007] After accessing historical wind data locally, the yaw coordination strategy is solved in the twin wind farm based on the historical wind data to build a benchmark coordination control library.

[0008] Connect to a remote weather forecasting center to access forecast wind data that falls within the forecast time window.

[0009] The predicted wind data is used to traverse the baseline collaborative control library, and the baseline collaborative yaw angle is output.

[0010] Starting from the baseline coordinated yaw angle, the yaw coordinated control optimization is performed in the twin wind farm based on the predicted wind condition data, and the predicted coordinated yaw angle is output.

[0011] Within the predicted time window, the power generation intervention of the offshore floating wind farm is carried out based on the predicted cooperative yaw angle.

[0012] In one feasible implementation, a digital twin of the offshore floating wind farm is constructed to obtain the twin wind farm, including:

[0013] The design parameters of the offshore floating wind farm were collected, and a wind farm control coupling model was established.

[0014] By integrating a CFD wake model into the wind farm control coupling model, the superposition effect of turbine wakes and the spatiotemporal difference effect are quantified, and the twin wind farm is output.

[0015] Based on the historical wind data, historical wind turbine yaw coordination data and historical power generation data are retrieved in a time-series synchronous manner.

[0016] The historical wind data, historical wind turbine yaw coordination data, and historical power generation data are used to calibrate the accuracy of the wake field distribution of the twin wind farm.

[0017] In one feasible implementation, after locally accessing historical wind condition data, a yaw coordination strategy is solved in the twin wind farm based on the historical wind condition data to construct a benchmark coordination control library, including:

[0018] Based on the interval discretization processing of the historical wind data, multiple sample discrete wind scenario scenarios are obtained.

[0019] With the goal of maximizing the overall power generation, the cooperative yaw angle of multiple samples in the twin wind farm is solved for the discrete wind condition scenarios.

[0020] The system associates and stores the discrete wind condition scenarios and the coordinated yaw angles of the multiple samples, and outputs the benchmark coordinated control library.

[0021] In one feasible implementation, the historical wind condition data is processed by interval discretization to obtain multiple sample discrete wind condition scenarios, including:

[0022] Discretized core parameter indicators are defined, wherein the discretized core parameter indicators consist of wind speed, wind direction, and wave height.

[0023] Define the index interval division rules for the discretized core parameter index.

[0024] Using the discretized core parameter indicators as constraints, multiple historical wind condition records corresponding to multiple historical acquisition nodes are retrieved from the historical wind condition data.

[0025] Based on the index interval division rules, the multiple historical wind condition records are divided into multiple wind condition scenario discrete matrices.

[0026] By aggregating the discrete matrices of the multiple wind condition scenarios, the multiple sample discrete wind condition scenarios are obtained.

[0027] In one feasible implementation, with the goal of maximizing the overall power generation, the twin wind farm solves for multiple sample coordinated yaw angles of the multiple sample discrete wind condition scenarios, including:

[0028] By using the first sample discrete wind condition scenario to call historical yaw features, multiple historical coordinated power generation values ​​at multiple historical coordinated yaw angles are obtained.

[0029] Based on the multiple historical coordinated power generation values, a benchmark coordinated yaw angle is selected from the multiple historical coordinated yaw angles.

[0030] A first sample discrete wind condition scenario and the benchmark cooperative yaw angle are loaded into the twin wind farm to initialize the wind turbine layout and floating platform dynamic parameters.

[0031] Starting from the aforementioned benchmark coordinated yaw angle, the yaw angle of the wind turbines in the twin wind farm is adjusted using the gradient descent method until the first sample coordinated yaw angle that maximizes the power generation of the entire farm is iteratively output.

[0032] In one feasible implementation, the predicted wind data is used to traverse the baseline cooperative control library to output the baseline cooperative yaw angle, including:

[0033] Based on the numerical inclusion relationship between the predicted wind condition data and the multiple sample discrete wind condition scenarios in the benchmark collaborative control library, the benchmark collaborative yaw angle is located at the multiple sample collaborative yaw angles.

[0034] The reference coordinated yaw angle is output to the twin wind farm for yaw coordinated control optimization.

[0035] In one feasible implementation, within the predicted time window, the power generation intervention of the offshore floating wind farm is performed based on the predicted cooperative yaw angle, including:

[0036] Call the real-time collaborative yaw angle.

[0037] Calculate the quantitative value of the collaborative difference between the real-time collaborative yaw angle and the predicted collaborative yaw angle.

[0038] If the quantified value of the collaborative difference is less than the preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm will continue with the real-time collaborative yaw angle.

[0039] If the quantified value of the collaborative difference is greater than the preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm is carried out according to the predicted collaborative yaw angle.

[0040] Secondly, the present invention also provides a wind turbine yaw coordinated control system for an offshore floating wind farm, wherein the wind turbine yaw coordinated control system for the offshore floating wind farm includes:

[0041] The twin construction module is used to build a digital twin of an offshore floating wind farm, thus obtaining a twin wind farm.

[0042] The benchmark library construction module is used to call historical wind condition data locally and then solve the yaw coordination strategy in the twin wind farm based on the historical wind condition data in order to construct a benchmark coordination control library.

[0043] The weather forecast access module is used to connect to a remote weather forecast center to access forecast wind condition data that falls within the forecast time window.

[0044] The yaw angle matching module is used to traverse the benchmark collaborative control library using the predicted wind condition data and output the benchmark collaborative yaw angle.

[0045] The control optimization module is used to perform yaw coordination control optimization in the twin wind farm based on the predicted wind condition data, starting from the benchmark coordinated yaw angle, and output the predicted coordinated yaw angle.

[0046] The power generation intervention module is used to intervene in the power generation of the offshore floating wind farm within the predicted time window based on the predicted cooperative yaw angle.

[0047] This invention discloses a method and system for coordinated yaw control of wind turbines in offshore floating wind farms, comprising: constructing a digital twin model corresponding to the target offshore floating wind farm to form a twin wind farm for control simulation; performing yaw coordinated control strategy solution in the twin wind farm based on locally accessed historical wind condition data, and constructing and storing a benchmark coordinated control strategy library; accessing a remote weather forecast service to obtain predicted wind condition data within the target prediction time window; inputting the predicted wind condition data into the benchmark coordinated control strategy library, performing control strategy matching, and outputting the corresponding benchmark coordinated yaw angle; using the benchmark coordinated yaw angle as the initial control condition, performing coordinated control optimization in the twin wind farm based on the predicted wind condition data to obtain the optimized predicted coordinated yaw angle; and implementing yaw control intervention on the offshore floating wind farm within the prediction time window based on the predicted coordinated yaw angle. The method and system for coordinated yaw control of wind turbines in offshore floating wind farms disclosed in this invention solves the technical problems of lacking a perspective on wind condition changes and historical patterns, having a single control dimension, and insufficient predictability, achieving the technical effects of forward-looking power generation intervention and improving power generation efficiency and stability. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the wind turbine yaw coordinated control method for offshore floating wind farms according to the present invention.

[0049] Figure 2 This is a schematic diagram of the wind turbine yaw coordinated control system for the offshore floating wind farm of the present invention.

[0050] Figure labeling: Twin construction module 11, benchmark library construction module 12, weather forecast access module 13, yaw angle matching module 14, control optimization module 15, power generation intervention module 16. Detailed Implementation

[0051] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them.

[0052] Example 1, as Figure 1 This is a flowchart illustrating the wind turbine yaw coordinated control method for an offshore floating wind farm according to the present invention. The wind turbine yaw coordinated control method for the offshore floating wind farm includes:

[0053] S100: Construct a digital twin of an offshore floating wind farm to obtain a twin wind farm.

[0054] Specifically, a digital twin of an offshore floating wind farm refers to a virtual simulation model and real-time synchronization mechanism built based on the structural, state, and environmental data of an actual floating wind farm. This model is used to simulate, predict, and provide feedback on the operating status of various equipment and environmental changes within the wind farm. The digital twin wind farm is a mapping of the real wind farm to the digital twin during operation, possessing the ability to synchronize data input, state evolution, and behavioral output with the real system.

[0055] The twin wind farms constructed through the above process can perform highly consistent dynamic response simulations and parameter predictions without direct intervention in the actual wind turbines, providing real-time visual feedback and predictive evaluation basis for the coordinated adjustment of wind turbine yaw angles.

[0056] In some embodiments, a digital twin of an offshore floating wind farm is constructed to obtain a twin wind farm, including:

[0057] Design parameters of the offshore floating wind farm are collected to establish a wind farm control coupling model. A CFD wake model is integrated into the wind farm control coupling model to quantify the superposition effect and spatiotemporal difference effect of the turbine cluster wake, and the twin wind farm is output. Based on the historical wind condition data, historical wind turbine yaw coordination data and historical power generation data are synchronously called in time sequence. The wake field distribution accuracy of the twin wind farm is calibrated using the historical wind condition data, historical wind turbine yaw coordination data and historical power generation data.

[0058] Specifically, a digital twin of an offshore floating wind farm is constructed. Based on the physical structure and operational characteristics of the actual wind farm, a virtual model system capable of dynamically simulating and providing feedback on the wind farm's state is generated through modeling and data synchronization. For example, this digital twin includes a wind farm control coupling model, a wake effect modeling module, and a data calibration module. The control coupling model simulates the operational control relationships between multiple floating wind turbines, the CFD wake model analyzes the spatial propagation and superposition effects of the turbine wake, and the wake field distribution accuracy calibration utilizes historical data to improve the accuracy of the twin model.

[0059] Specifically, firstly, design parameters for offshore floating wind farms are collected, including but not limited to the hub height of the wind turbines, blade length, platform six-degree-of-freedom response characteristics, floating system mooring parameters, cable layout information, and wind turbine control strategy parameters. Then, a wind farm control coupling model considering attitude disturbances, load transfer, and control interaction is established. Next, a CFD (Computational Fluid Dynamics) wake model is integrated into this coupling model to simulate the wake generated by single and multiple wind turbines under complex wind conditions. In this process, the spatiotemporal attenuation, lateral drift, and superposition coupling effects of the wake need to be considered. This allows for the quantification of the wake interference intensity and mutual influence relationship between the wind turbines in the cluster, and the output of a twin wind farm with wake correlation state mapping capability.

[0060] In addition, to enhance the accuracy of model prediction, historical wind data (such as wind speed and direction time series curves), historical wind turbine yaw coordination control data, and historical power generation data are further retrieved in time series. Based on this, deviation analysis and optimization adjustments are made to the CFD wake simulation results to complete the accuracy calibration of the wake field distribution of the twin wind farm.

[0061] The digital twin wind farm constructed through the above steps can not only accurately reproduce the yaw response and power generation performance of offshore floating wind farms under specific wind conditions, but also significantly improve the quantification capability of wake disturbance effects and the response prediction accuracy of inter-turbine collaborative control. Compared with traditional static models, this twin system has stronger dynamic consistency and spatiotemporal fitting capabilities, enabling it to support the simulation verification, optimization, and rolling updates of wind turbine yaw collaborative control strategies before deployment, thereby improving the overall power generation efficiency and energy dispatch flexibility of offshore wind farms.

[0062] S200: After locally calling historical wind data, the system solves the yaw coordination strategy in the twin wind farm based on the historical wind data to build a benchmark coordination control library.

[0063] Specifically, pre-saved wind condition information, including but not limited to historical data records such as wind speed, wind direction, and wind frequency, is extracted from the local storage system. Then, through specific algorithms or models, control strategies that can optimize the yaw coordination of wind turbines are found in the digital twin wind farm based on the historical wind condition data, so as to achieve efficient coordinated operation between wind turbines.

[0064] In some embodiments, after locally retrieving historical wind condition data, a yaw coordination strategy is solved in the twin wind farm based on the historical wind condition data to construct a benchmark coordination control library, including:

[0065] Based on the interval discretization processing of the historical wind condition data, multiple sample discrete wind condition scenarios are obtained; with the goal of maximizing the overall power generation, multiple sample coordinated yaw angles of the multiple sample discrete wind condition scenarios are solved in the twin wind farm; the multiple sample discrete wind condition scenarios and multiple sample coordinated yaw angles are associated and stored, and the benchmark coordinated control library is output.

[0066] Specifically, interval discretization involves dividing the originally continuously changing historical wind data (such as wind speed and wind direction) into several representative discrete intervals according to a certain interval division method, thereby forming a limited number of typical wind scenario scenarios, which facilitates subsequent batch optimization and strategy library construction.

[0067] Specifically, cooperative yaw angle refers to the optimal combination of yaw angles determined by all wind turbines in a wind farm through collaborative optimization under specific wind conditions. This combination of angles aims to improve overall power generation or reduce load, rather than the independent optimality of a single turbine. The benchmark cooperative control library maps different wind conditions to corresponding optimal cooperative yaw angles through offline simulation and optimization, and stores the resulting control strategy database for quick lookup and retrieval during actual operation.

[0068] Specifically, firstly, based on predetermined discretization rules (such as discretization based on wind speed level ranges), the historical wind speed, wind direction, and other wind condition data collected from wind farms in recent years are discretized. For example, wind speed is divided into ranges such as 3–5 m / s, 5–7 m / s, and 7–9 m / s, and wind direction is divided into ranges such as 0–30° and 30–60°, and these are combined to form several typical wind condition scenario samples (such as wind speed of 5–7 m / s and wind direction of 30–60°).

[0069] Then, using the twin wind farm model, collaborative optimization is performed on each discrete wind condition scenario. That is, under each discretized wind condition scenario, with the goal of "maximizing the power generation of the entire field", intelligent optimization methods such as genetic algorithms and particle swarm optimization are used to solve the optimal collaborative yaw angle combination of all wind turbines in the wind farm.

[0070] Furthermore, the aforementioned discrete wind condition scenarios and their corresponding sample coordinated yaw angles are associated and stored to form a benchmark coordinated control library. For example, the benchmark coordinated control library includes the following: the yaw angle set corresponding to scenario 1 (wind speed 5-10m / s, wind direction 45-90°, wave height 1-2m) is {Wind turbine A: 15°, Wind turbine B: 20°, Wind turbine C: 10°}; the yaw angle set corresponding to scenario 2 (wind speed 10-15m / s, wind direction 90-135°, wave height 2-3m) is {Wind turbine A: 25°, Wind turbine B: 30°, Wind turbine C: 20°}, etc.

[0071] Through the above process, the standardization and efficiency of wind farm operation strategies can be achieved. First, the benchmark collaborative control library established based on historical data and simulation optimization can significantly improve the overall power generation efficiency of the wind farm and reduce wake losses. Second, the pre-construction of the control library greatly reduces the online computational burden, enabling the wind farm to respond quickly to changes in wind conditions during actual operation and improving its adaptability. Furthermore, this method has good scalability and maintainability; the control library can be updated periodically based on newly collected historical data to continuously optimize the wind farm's operation strategy.

[0072] In some implementations, the historical wind data is processed by interval discretization to obtain multiple sample discrete wind scenario scenarios, including:

[0073] Discretized core parameter indicators are defined, wherein the discretized core parameter indicators consist of wind speed, wind direction, and wave height indicators; the indicator interval division rules of the discretized core parameter indicators are defined; using the discretized core parameter indicators as constraints, multiple historical wind condition records corresponding to multiple historical acquisition nodes are retrieved from the historical wind condition data; the multiple historical wind condition records are divided into multiple wind condition scene discrete matrices according to the indicator interval division rules; the multiple wind condition scene discrete matrices are aggregated to obtain the multiple sample discrete wind condition scenes.

[0074] Specifically, the discretized core parameter index refers to the key physical quantities selected when characterizing and classifying wind conditions. These parameters directly affect the operating status and power generation efficiency of wind turbines, including wind speed, wind direction, and wave height.

[0075] Specifically, the index interval division rule refers to the specific method for dividing the above parameters into intervals, such as equal-interval division, equal-frequency division, or adaptive division based on statistical distribution. Historical data acquisition nodes refer to various sensors or measuring points deployed within the wind farm for collecting wind condition data. The wind condition scenario discrete matrix refers to organizing historical wind condition data in matrix form within a specific parameter interval for subsequent analysis and modeling.

[0076] Specifically, first, discretize core parameter indicators, such as wind speed (m / s), wind direction (°), and wave height (m). Then, define interval division rules for each indicator. For example, wind speed can be set to three intervals: 0-4, 4-8, and 8-12 m / s; wind direction can be set to 0-60°, 60-120°, and 120-180°; and wave height can be set to 0-1m, 1-2m, and 2-3m. Next, extract historical wind speed, wind direction, and wave height records for the corresponding time periods from all historical data collection nodes. Based on the interval division rules, divide all historical wind condition records into different wind condition scenario discrete matrices. For example, if at a certain moment, the wind speed at node A is 6 m / s, the wind direction is 75°, and the wave height is 1.2m, then this data is classified into the scenario of "wind speed 4-8 m / s, wind direction 60-120°, wave height 1-2m".

[0077] Furthermore, similar processing is performed on all data collection nodes and times to obtain multiple sets of discrete wind condition scene matrices. These matrices are then aggregated according to interval combinations to obtain a series of representative sample discrete wind condition scenes. Each scene represents wind condition characteristics within a certain range. For example, a sample discrete wind condition scene may include all historical records with wind speeds between 5-10 m / s, wind directions between 45-90°, and wave heights between 1-2 m.

[0078] Through the above process, the historical wind data was structured and standardized, effectively extracting typical wind condition scenarios for wind farm operation. The interval discretization method based on multiple core parameters considers not only the direct impact of wind speed and direction on wind turbine power generation but also incorporates environmental factors such as wave height, making the sample scenarios more representative and applicable. Aggregating a large amount of historical data reduces the interference of occasional extreme wind conditions on subsequent optimization, improving the statistical reliability of the samples.

[0079] In some implementations, with the goal of maximizing the overall power generation, the twin wind farm solves for multiple sample coordinated yaw angles of the multiple sample discrete wind condition scenarios, including:

[0080] Historical yaw features are retrieved using a first sample discrete wind condition scenario to obtain multiple historical coordinated power generation values ​​for multiple historical coordinated yaw angles. Based on these multiple historical coordinated power generation values, a benchmark coordinated yaw angle is selected from the multiple historical coordinated yaw angles. The first sample discrete wind condition scenario and the benchmark coordinated yaw angle are loaded into the twin wind farm to initialize the wind turbine layout and floating platform dynamic parameters. Starting from the benchmark coordinated yaw angle, the yaw angle of the wind turbines is adjusted in the twin wind farm using a gradient descent method until the first sample coordinated yaw angle that maximizes the overall power generation of the farm is iteratively output.

[0081] Specifically, historical coordinated yaw angles refer to the specific yaw settings of each wind turbine under historical wind conditions. Historical coordinated power generation refers to the actual power generation under these yaw settings. The benchmark coordinated yaw angle is the optimal combination of angles for power generation selected from historical yaw angles by comparing historical coordinated power generation.

[0082] Specifically, initializing the wind turbine layout and floating platform dynamic parameters refers to setting the spatial arrangement of the wind turbines and the dynamic state of the floating foundation in the twin wind farm simulation environment.

[0083] Specifically, firstly, a scenario is randomly selected as the first sample discrete wind condition scenario, and corresponding historical databases are retrieved for all yaw angle combinations and their power generation records under similar wind conditions. For example, for the scenario of "wind speed 4-8 m / s, wind direction 60-120°, wave height 1-2 m", the yaw angles (such as [8°, 10°, 9°], [6°, 11°, 7°], etc.) and their corresponding total power generation used by the corresponding wind turbines in historical operation are retrieved. Then, from these historical coordinated yaw angles, the combination with the highest coordinated power generation is selected as the benchmark coordinated yaw angle for this scenario.

[0084] Furthermore, in the twin wind farm simulation model, the spatial position of the wind turbine and the dynamic parameters of the floating platform are initialized based on the sample wind condition scenario and the benchmark coordinated yaw angle. Then, using the benchmark coordinated yaw angle as the initial point, the wind turbine yaw angle is fine-tuned using the gradient descent method. After each adjustment, the power generation of the entire field is calculated through simulation until it converges to the maximum power generation, and finally the optimal coordinated yaw angle under this scenario is output.

[0085] Through the above process, historical operating experience and simulation optimization capabilities can be fully integrated to further improve power generation performance. Starting with the historically optimal yaw angle helps accelerate the convergence speed of the gradient descent method, reduce invalid searches, and improve optimization efficiency. Meanwhile, the introduction of twin wind farms ensures the physical feasibility and engineering applicability of the optimization results.

[0086] S300: Connects to a remote weather forecasting center to access forecast wind data that falls within the forecast time window.

[0087] Specifically, a remote weather forecasting center refers to an external weather forecasting service center that provides real-time weather forecast data. For example, predicted wind condition data includes information such as wind speed, wind direction, and air pressure for a future period. The forecast time window is a preset time range that can be determined based on the actual needs of the wind farm and the accuracy of the weather forecast, such as 24 hours, 48 ​​hours, or 72 hours in advance, to ensure that the acquired weather forecast data has high reference value. Predicted wind condition data, on the other hand, is the forecast information about future wind speed, wind direction, and other wind conditions provided by the weather forecasting center. This data will be used for the forward-looking control of the wind farm.

[0088] S400: Use the predicted wind data to traverse the benchmark collaborative control library and output the benchmark collaborative yaw angle.

[0089] Specifically, the predicted wind condition data is a continuous data stream composed of information such as wind speed and direction for a future period of time, obtained from the wind farm's environmental sensing system or external meteorological models. Through traversal operations, the predicted wind conditions can be matched and identified with scenarios in the control library, and the yaw angle corresponding to the closest scenario is output as the benchmark coordinated yaw angle. This provides a better starting point for subsequent control optimization, thereby improving the decision-making efficiency of the current wind farm's yaw control.

[0090] In some embodiments, the predicted wind data is used to traverse the baseline cooperative control library to output a baseline cooperative yaw angle, including:

[0091] Based on the numerical inclusion relationship between the predicted wind condition data and the multiple sample discrete wind condition scenarios in the benchmark collaborative control library, the benchmark collaborative yaw angle is located at the multiple sample collaborative yaw angles; the benchmark collaborative yaw angle is then called and output to the twin wind farm for yaw collaborative control optimization.

[0092] Specifically, first, predicted wind condition data is acquired, including wind speed, wind direction, and height profile information for each sub-time period within the prediction timeframe. Then, the predicted wind condition data is compared with multiple discrete wind condition scenarios in a benchmark collaborative control library using numerical inclusion or similarity comparisons to identify the nearest neighboring scenario in the library corresponding to the predicted wind condition data. For example, Euclidean distance, wind direction interval envelope, and wind speed vector similarity can be used for this determination.

[0093] Furthermore, after identifying the best matching sample, the yaw angle in the set of coordinated yaw angles corresponding to that sample is taken as the benchmark coordinated yaw angle, which is the wind turbine group coordinated yaw control strategy recommended under the current wind conditions. Finally, the benchmark coordinated yaw angle is input into the digital twin wind farm control model as the initial input or search guide for yaw control optimization, and further subsequent control strategy iteration and scheduling optimization are carried out.

[0094] The above-mentioned sample matching and benchmark yaw angle output process based on predicted wind data enables the control system to generate initial values ​​for collaborative control with global reference value in advance before future wind changes occur, effectively shortening the optimization delay and avoiding entering the invalid solution region. Through the high-quality sample strategy pre-embedded in the benchmark collaborative control library, the control of minimizing wake effects can be achieved while ensuring the overall output power stability of the wind farm, thereby improving the group power generation efficiency and wind turbine life utilization rate.

[0095] S500: Starting from the reference coordinated yaw angle, perform yaw coordinated control optimization in the twin wind farm based on the predicted wind condition data, and output the predicted coordinated yaw angle.

[0096] Specifically, in actual operation, the yaw strategy of the wind farm needs to be dynamically adjusted according to future wind conditions. First, the baseline coordinated yaw angle obtained by optimizing the aforementioned historical data is used as the initial solution, and the latest predicted wind condition data (such as wind speed of 7 m / s, wind direction of 90°, and wave height of 1.5 m in the next hour) is loaded into the twin wind farm simulation platform. Then, with the predicted wind conditions as constraints, numerical optimization methods (such as gradient descent or genetic algorithms) are used to coordinately optimize the yaw angle of each wind turbine.

[0097] Specifically, during the optimization process, the simulation platform calculates the impact of different yaw combinations on the overall power generation in real time, gradually adjusts the yaw angle until the power generation converges to the maximum value or reaches the preset number of optimization steps, selects the yaw angle combination that maximizes the overall power generation, and finally outputs the optimal coordinated yaw angle under the predicted wind conditions, which serves as the actual yaw setting for each unit in the wind farm in the next control cycle and is sent to the wind farm control system in real time.

[0098] Through the above process, the wind farm achieves forward-looking adaptive control of future wind conditions, dynamically adjusting the yaw strategy of each turbine based on the predicted environment. Starting with historical optimization results significantly shortens the convergence time of the optimization algorithm and improves the response speed and accuracy of the control strategy. High-precision simulation of the twin wind farm ensures the physical feasibility and engineering practicality of the optimization results.

[0099] S600: Within the predicted time window, the power generation intervention of the offshore floating wind farm is carried out based on the predicted cooperative yaw angle.

[0100] In some embodiments, within the prediction time window, the power generation intervention of the offshore floating wind farm is performed based on the predicted cooperative yaw angle, including:

[0101] The system calls the real-time coordinated yaw angle; calculates the coordinated difference quantification value between the real-time coordinated yaw angle and the predicted coordinated yaw angle; if the coordinated difference quantification value is less than a preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm continues with the real-time coordinated yaw angle; if the coordinated difference quantification value is greater than the preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm is carried out according to the predicted coordinated yaw angle.

[0102] Specifically, the real-time coordinated yaw angle refers to the combination of yaw angles currently applied to each wind turbine in the wind farm. The coordinated difference quantification value refers to the overall deviation between the real-time coordinated yaw angle and the predicted coordinated yaw angle, quantitatively measured by mathematical methods (such as Euclidean distance, weighted absolute sum, etc.). The preset difference threshold is a pre-set numerical limit used to determine whether the current control strategy needs to be switched.

[0103] Specifically, within the prediction time window (e.g., the next 10 minutes), the real-time coordinated yaw angles of each wind turbine are first obtained (e.g., [10.0°, 11.0°, 9.0°]), and compared with the previously optimized predicted coordinated yaw angles (e.g., [9.8°, 11.2°, 8.4°]). A coordinated difference quantification method is used, for example, to calculate the Euclidean distance between the two sets of yaw angles.

[0104] Then, the quantified value of the collaborative difference is compared with the preset difference threshold. If the calculated quantified value of the collaborative difference is less than the preset difference threshold, the current real-time collaborative yaw angle is considered to be highly consistent with the predicted collaborative yaw angle, and no adjustment is needed. The current real-time collaborative yaw angle is continued for wind farm power generation intervention to avoid mechanical losses and control system burden caused by frequent adjustments. If the quantified value of the collaborative difference is greater than the preset difference threshold, it indicates that there is a significant difference between the current actual yaw state and the optimal predicted yaw state. Based on the predicted collaborative yaw angle, the yaw angle of each wind turbine will be adjusted accordingly, and a new power generation intervention strategy will be implemented to improve the overall power generation performance of the wind farm.

[0105] Through the above process, dynamic adaptive adjustment of the wind farm's yaw control is achieved. This ensures the system's sensitive response to changes in wind conditions while avoiding frequent switching of control commands due to minor deviations, thus extending equipment lifespan and improving operational stability. In other words, these steps effectively reduce the energy consumption and maintenance costs of the control system while maximizing power generation efficiency, thereby improving the economics and reliability of the wind farm.

[0106] In summary, the wind turbine yaw coordinated control method for offshore floating wind farms provided by this invention has the following technical effects:

[0107] By constructing a digital twin model corresponding to the target offshore floating wind farm, a twin wind farm for control simulation is formed. Based on locally accessed historical wind condition data, a yaw cooperative control strategy is solved in the twin wind farm, and a benchmark cooperative control strategy library is constructed and stored. A remote weather forecast service is accessed to obtain predicted wind condition data within the target prediction time window. The predicted wind condition data is input into the benchmark cooperative control strategy library for control strategy matching, and the corresponding benchmark cooperative yaw angle is output. Using the benchmark cooperative yaw angle as the initial control condition, cooperative control optimization is performed in the twin wind farm based on the predicted wind condition data to obtain the optimized predicted cooperative yaw angle. Within the prediction time window, based on the predicted cooperative yaw angle, yaw control intervention is implemented on the offshore floating wind farm, thereby achieving the technical effects of forward-looking power generation intervention and improving power generation efficiency and stability.

[0108] Example 2, as Figure 2 This is a schematic diagram of the wind turbine yaw coordinated control system for an offshore floating wind farm according to the present invention. For example, Figure 1 A flowchart illustrating the wind turbine yaw coordinated control method for offshore floating wind farms according to the present invention can be seen as follows: Figure 2 The structure shown is implemented.

[0109] Based on the same concept as the wind turbine yaw coordinated control method for offshore floating wind farms in the above embodiments, the present invention also provides a wind turbine yaw coordinated control system for offshore floating wind farms, comprising:

[0110] The twin construction module 11 is used to construct a digital twin of an offshore floating wind farm to obtain a twin wind farm.

[0111] The benchmark library construction module 12 is used to call historical wind data locally and then solve the yaw coordination strategy in the twin wind farm based on the historical wind data in order to construct a benchmark coordination control library.

[0112] The weather forecast access module 13 is used to access a remote weather forecast center to call up the forecast wind condition data that falls within the forecast time window.

[0113] The yaw angle matching module 14 is used to traverse the benchmark collaborative control library using the predicted wind condition data and output the benchmark collaborative yaw angle.

[0114] The control optimization module 15 is used to perform yaw coordination control optimization in the twin wind farm based on the predicted wind condition data, starting from the benchmark coordinated yaw angle, and output the predicted coordinated yaw angle.

[0115] The power generation intervention module 16 is used to intervene in the power generation of the offshore floating wind farm according to the predicted cooperative yaw angle within the predicted time window.

[0116] In some embodiments, the twin building module 11 includes:

[0117] The design parameter acquisition and coupling model establishment unit is used to acquire the design parameters of the offshore floating wind farm and establish the wind farm control coupling model.

[0118] The CFD wake model integration and twin wind farm output unit is used to integrate the CFD wake model into the wind farm control coupling model to quantify the superposition effect and spatiotemporal difference effect of the turbine cluster wake, and output the twin wind farm.

[0119] The historical data time-series synchronization calling unit is used to call historical wind turbine yaw coordination data and historical power generation data in time sequence according to the historical wind condition data.

[0120] The wake field distribution accuracy calibration unit is used to calibrate the wake field distribution accuracy of the twin wind farm using the historical wind condition data, historical wind turbine yaw coordination data, and historical power generation data.

[0121] In some embodiments, the benchmark library construction module 12 includes:

[0122] The historical wind condition data discretization processing unit is used to perform interval discretization processing on the historical wind condition data to obtain multiple sample discrete wind condition scenarios.

[0123] The sample collaborative yaw angle solving unit is used to solve multiple sample collaborative yaw angles in the twin wind farm with the goal of maximizing the overall power generation.

[0124] The benchmark collaborative control library output unit is used to associate and store the multiple sample discrete wind condition scenarios and the multiple sample collaborative yaw angles, and output the benchmark collaborative control library.

[0125] In some implementations, the execution steps of the historical wind data discretization processing unit in the benchmark library construction module 12 include:

[0126] Discretized core parameter indicators are defined, wherein the discretized core parameter indicators consist of wind speed, wind direction, and wave height.

[0127] Define the index interval division rules for the discretized core parameter index.

[0128] Using the discretized core parameter indicators as constraints, multiple historical wind condition records corresponding to multiple historical acquisition nodes are retrieved from the historical wind condition data.

[0129] Based on the index interval division rules, the multiple historical wind condition records are divided into multiple wind condition scenario discrete matrices.

[0130] By aggregating the discrete matrices of the multiple wind condition scenarios, the multiple sample discrete wind condition scenarios are obtained.

[0131] In some implementations, the execution steps of the sample cooperative yaw angle solving unit in the benchmark library construction module 12 include:

[0132] By using the first sample discrete wind condition scenario to call historical yaw features, multiple historical coordinated power generation values ​​at multiple historical coordinated yaw angles are obtained.

[0133] Based on the multiple historical coordinated power generation values, a benchmark coordinated yaw angle is selected from the multiple historical coordinated yaw angles.

[0134] A first sample discrete wind condition scenario and the benchmark cooperative yaw angle are loaded into the twin wind farm to initialize the wind turbine layout and floating platform dynamic parameters.

[0135] Starting from the aforementioned benchmark coordinated yaw angle, the yaw angle of the wind turbines in the twin wind farm is adjusted using the gradient descent method until the first sample coordinated yaw angle that maximizes the power generation of the entire farm is iteratively output.

[0136] In some embodiments, the yaw angle matching module 14 includes:

[0137] The benchmark cooperative yaw angle positioning unit is used to locate the benchmark cooperative yaw angle based on the numerical inclusion relationship between the predicted wind condition data and the multiple sample discrete wind condition scenarios in the benchmark cooperative control library.

[0138] The yaw cooperative control optimization invocation unit is used to invoke the output of the reference cooperative yaw angle to the twin wind farm for yaw cooperative control optimization.

[0139] In some embodiments, the power generation intervention module 16 includes:

[0140] The real-time collaborative yaw angle calling unit is used to call the real-time collaborative yaw angle.

[0141] The collaborative difference quantification unit is used to calculate the collaborative difference quantification value between the real-time collaborative yaw angle and the predicted collaborative yaw angle.

[0142] The power generation intervention decision unit is used to intervene in the power generation of the offshore floating wind farm by continuing the real-time coordinated yaw angle within the prediction time window if the coordinated difference quantification value is less than the preset difference threshold; and to intervene in the power generation of the offshore floating wind farm by adjusting the predicted coordinated yaw angle within the prediction time window if the coordinated difference quantification value is greater than the preset difference threshold.

[0143] It should be understood that the focus of the embodiments mentioned in this specification is their difference from other embodiments. The specific embodiments in the aforementioned Embodiment 1 are also applicable to the wind turbine yaw coordinated control system of the offshore floating wind farm described in Embodiment 2. For the sake of brevity, they will not be further elaborated here.

[0144] It should be understood that the embodiments disclosed in this invention and the above description enable those skilled in the art to implement this invention. However, this invention is not limited to the embodiments mentioned above. It should be understood that those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention, and should all be included within the protection scope of this invention.

Claims

1. A method for coordinated yaw control of wind turbines in offshore floating wind farms, characterized in that, include: Construct a digital twin of an offshore floating wind farm to obtain a twin wind farm; After locally accessing historical wind data, the yaw coordination strategy is solved in the twin wind farm based on the historical wind data to build a benchmark coordination control library. Connect to a remote weather forecasting center to access forecast wind data that falls within the forecast time window; The predicted wind data is used to traverse the benchmark cooperative control library and the benchmark cooperative yaw angle is output. Starting from the baseline cooperative yaw angle, the yaw cooperative control optimization is performed in the twin wind farm based on the predicted wind condition data, and the predicted cooperative yaw angle is output. Within the predicted time window, the power generation intervention of the offshore floating wind farm is carried out based on the predicted cooperative yaw angle. Among them, constructing a digital twin of an offshore floating wind farm results in a twin wind farm, including: The design parameters of the offshore floating wind farm were collected, and a wind farm control coupling model was established. By integrating a CFD wake model into the wind farm control coupling model, the superposition effect of turbine wakes and the spatiotemporal difference effect are quantified, and the twin wind farm is output. Based on the historical wind data, historical wind turbine yaw coordination data and historical power generation data are retrieved in a time-series synchronous manner. The historical wind condition data, historical wind turbine yaw coordination data, and historical power generation data are used to calibrate the wake field distribution accuracy of the twin wind farm. Specifically, after locally accessing historical wind data, the system solves for a yaw coordination strategy in the twin wind farm based on this historical wind data to construct a benchmark coordination control library, including: Based on the interval discretization processing of the historical wind data, multiple sample discrete wind scenario is obtained; With the goal of maximizing the overall power generation, the cooperative yaw angle of multiple samples in the multiple discrete wind condition scenarios is solved in the twin wind farm. The system associates and stores the discrete wind condition scenarios and the coordinated yaw angles of the multiple samples, and outputs the benchmark coordinated control library. Among them, based on the interval discretization processing of the historical wind condition data, multiple sample discrete wind condition scenarios are obtained, including: Discretized core parameter indicators are defined, wherein the discretized core parameter indicators consist of wind speed, wind direction, and wave height. Define the index interval division rules for the discretized core parameter index; Using the discretized core parameter indicators as constraints, multiple historical wind condition records corresponding to multiple historical acquisition nodes are retrieved from the historical wind condition data; Based on the index interval division rules, the multiple historical wind condition records are divided into multiple wind condition scenario discrete matrices; By aggregating the discrete matrices of the multiple wind condition scenarios, the multiple sample discrete wind condition scenarios are obtained.

2. The wind turbine yaw coordinated control method for offshore floating wind farms as described in claim 1, characterized in that, With the goal of maximizing overall power generation, the cooperative yaw angles of multiple samples in the twin wind farm are solved for the discrete wind condition scenarios, including: By using the first sample discrete wind condition scenario to call historical yaw features, multiple historical coordinated power generation values ​​at multiple historical coordinated yaw angles are obtained; Based on the multiple historical coordinated power generation values, a benchmark coordinated yaw angle is selected from the multiple historical coordinated yaw angles; In the twin wind farm, a first sample discrete wind condition scenario and the benchmark cooperative yaw angle are loaded to initialize the wind turbine layout and floating platform dynamic parameters. Starting from the aforementioned benchmark coordinated yaw angle, the yaw angle of the wind turbines in the twin wind farm is adjusted using the gradient descent method until the first sample coordinated yaw angle that maximizes the power generation of the entire farm is iteratively output.

3. The wind turbine yaw coordinated control method for offshore floating wind farms as described in claim 1, characterized in that, The predicted wind data is used to traverse the baseline cooperative control library to output the baseline cooperative yaw angle, including: Based on the numerical inclusion relationship between the predicted wind condition data and the multiple sample discrete wind condition scenarios in the benchmark collaborative control library, the benchmark collaborative yaw angle is located at the multiple sample collaborative yaw angles. The reference coordinated yaw angle is output to the twin wind farm for yaw coordinated control optimization.

4. The wind turbine yaw coordinated control method for offshore floating wind farms as described in claim 1, characterized in that, Within the predicted time window, the power generation intervention of the offshore floating wind farm is performed based on the predicted cooperative yaw angle, including: Call the real-time collaborative yaw angle; Calculate the quantitative value of the cooperative difference between the real-time cooperative yaw angle and the predicted cooperative yaw angle; If the quantified value of the collaborative difference is less than the preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm will continue with the real-time collaborative yaw angle. If the quantified value of the collaborative difference is greater than the preset difference threshold, then within the prediction time window, the power generation intervention of the offshore floating wind farm is carried out according to the predicted collaborative yaw angle.

5. A wind turbine yaw coordinated control system for offshore floating wind farms, characterized in that, The wind turbine yaw coordinated control method for implementing any one of claims 1 to 4 of the offshore floating wind farm includes: The twin construction module is used to construct a digital twin of an offshore floating wind farm, thus obtaining a twin wind farm. The benchmark library construction module is used to call historical wind condition data locally and then solve the yaw coordination strategy in the twin wind farm based on the historical wind condition data in order to construct a benchmark coordination control library. The weather forecast access module is used to connect to a remote weather forecast center to retrieve forecast wind condition data that falls within the forecast time window; The yaw angle matching module is used to traverse the benchmark collaborative control library using the predicted wind condition data and output the benchmark collaborative yaw angle. The control optimization module is used to perform yaw coordination control optimization in the twin wind farm based on the predicted wind condition data, starting from the benchmark coordinated yaw angle, and output the predicted coordinated yaw angle. The power generation intervention module is used to intervene in the power generation of the offshore floating wind farm within the predicted time window based on the predicted cooperative yaw angle.

Images