Real-time Hybrid Simulation Test Method and System for Fixed Offshore Wind Turbine Systems

Abstract

This invention provides a method and system for real-time hybrid simulation testing of a fixed offshore wind turbine system. The method includes: reducing the (i-1)th wind load parameter according to the wind parameter ratio, and using the reduced (i-1)th wind load parameter to drive a complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model; collecting the i-th motion characteristics of the offshore wind turbine system model under the influence of the i-th simulated natural conditions; expanding the i-th motion characteristics according to the characteristic ratio relationship to obtain the i-th expanded motion characteristics; processing the i-th expanded motion characteristics using a target neural network to obtain the i-th wind load parameter; and the target neural network characterizing the nonlinear mapping relationship between the wind load parameters of the actual offshore wind turbine system and the corresponding motion characteristics.

Description

Technical Field

[0001] This invention relates to the fields of offshore wind turbine systems and artificial intelligence technology, specifically to a real-time hybrid simulation test method and system for a stationary offshore wind turbine system. Background Technology

[0002] Fixed offshore wind turbine systems are generally affected by complex natural conditions such as wind, waves, currents, and earthquakes. Research on hybrid simulation testing technology for offshore wind turbine systems is still limited to numerical experiments and small-scale physical experiments. Due to limitations in testing conditions, it is difficult to accurately simulate the motion characteristics of fixed offshore wind turbine systems in response to natural conditions. Summary of the Invention

[0003] In view of the above problems, the present invention provides a method and system for real-time hybrid simulation test of a stationary offshore wind turbine system.

[0004] According to a first aspect of the present invention, a method for real-time hybrid simulation testing of a fixed offshore wind turbine system is provided, comprising: scaling down an i-1th wind load parameter according to a wind parameter ratio, and using the scaled-down i-1th wind load parameter to drive a complex boundary coordination loading device to apply an i-th wind load to an offshore wind turbine system model; the offshore wind turbine system model is obtained by scaling down an actual offshore wind turbine system; the wind parameter ratio characterizes the proportional relationship between the wind load of the offshore wind turbine system model and the wind load of the actual offshore wind turbine system; i is a positive integer greater than 1; and collecting data from the offshore wind turbine system model under the i-th simulated natural conditions. Motion characteristics; the offshore wind turbine system model includes saturated soil, a foundation structure set on the saturated soil, and a wind turbine structure set on the foundation structure; the i-th simulated natural condition includes the i-th wind load; the i-th motion characteristic is expanded according to the characteristic proportional relationship to obtain the i-th expanded motion characteristic; the characteristic proportional relationship represents the proportional relationship between the motion characteristics of the actual offshore wind turbine system and the motion characteristics of the offshore wind turbine system model; the i-th expanded motion characteristic is processed using a target neural network to obtain the i-th wind load parameter; the target neural network represents the nonlinear mapping relationship between the wind load parameter of the actual offshore wind turbine system and the motion characteristic corresponding to the wind load parameter.

[0005] According to a second aspect of the present invention, a real-time hybrid simulation test system for a fixed offshore wind turbine system is provided, comprising an offshore wind turbine system model, a complex boundary coordination loading device, and a controller. The offshore wind turbine system model includes saturated soil, a foundation structure situated on the saturated soil, and a wind turbine structure situated on the foundation structure; the offshore wind turbine system model is obtained by scaling down an actual offshore wind turbine system. The controller is used to execute the aforementioned hybrid simulation test method.

[0006] According to an embodiment of the present invention, the motion characteristics of an offshore wind turbine system model set on saturated soil under wind load are collected, then amplified according to the characteristic ratio, and the amplified motion characteristics are input into a target neural network. Based on this, even if the foundation structure is set on saturated soil that has been damaged by seawater immersion, causing the offshore wind turbine system model to undergo complex motion under wind load due to the loose structure of the saturated soil, resulting in a non-linear relationship between wind load parameters and motion characteristics, the model parameters of the target neural network of the present invention can still converge during the fitting process. Therefore, the target neural network can still quickly and accurately output the corresponding wind load parameters based on the motion characteristics of this complex motion. Furthermore, in each round of simulation experiments, the wind load parameters output by the target neural network can be reduced, and the reduced wind load can be used to drive a complex boundary coordination loading device to apply accurate wind loads to the offshore wind turbine system model in a timely manner. This achieves real-time and accurate simulation of the motion characteristics of a fixed offshore wind turbine system under natural conditions. Furthermore, by using the reduced wind load parameters to drive the complex boundary coordination loading device to test the offshore wind turbine system model, which includes saturated soil, foundation structure, and wind turbine structure, the problem of scale inconsistency between the actual offshore wind turbine system and the offshore wind turbine system model is avoided, which is caused by the offshore wind turbine system model not accurately replicating the structure of the actual offshore wind turbine system. Attached Figure Description

[0007] The above-mentioned contents, other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the present invention with reference to the accompanying drawings, which will be described in conjunction with the drawings.

[0008] Figure 1 A flowchart of a real-time hybrid simulation test method for a stationary offshore wind turbine system according to an embodiment of the present invention is shown;

[0009] Figure 2 A flowchart of the training method for the target neural network in the real-time hybrid simulation test method for a fixed offshore wind turbine system according to an embodiment of the present invention is shown;

[0010] Figure 3 A flowchart illustrating the application of wind load according to an embodiment of the present invention is shown;

[0011] Figure 4 A schematic diagram of the operation of a real-time hybrid simulation test system for a stationary offshore wind turbine system according to an embodiment of the present invention is shown. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0013] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0014] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0015] In the description of this invention, it should be understood that the terms "longitudinal", "length", "circumferential", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the subsystem or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0016] Similarly, to simplify the invention and aid in understanding one or more aspects, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof in the above description of exemplary embodiments of the invention. The use of terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples" indicates that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0017] In simulation tests of stationary offshore wind turbine systems, challenges exist in reproducing turbulent wind loads, as well as Reynolds-Froude scaling inconsistencies, structure-fluid scaling inconsistencies, and structure-saturated soil scaling inconsistencies between the actual offshore wind turbine system and its model. Research on hybrid simulation testing techniques for offshore wind turbine systems remains at the stage of numerical experiments and small-scale physical experiments. Due to limitations in experimental conditions, it is difficult to simultaneously consider the combined effects of wind, wave, current, and earthquakes, and to accurately reproduce the failure mechanism of the prototype structure (i.e., the actual offshore wind turbine system) under multi-load coupling. Furthermore, many hybrid simulation test methods for offshore wind turbine structures rely on numerical substructures calculated using simulation software, resulting in slow calculation speeds. Especially when structural stiffness degradation and / or soil plastic deformation occur, the numerical substructures are prone to non-convergence, making real-time data interaction between the numerical and experimental substructures difficult.

[0018] In view of this, embodiments of the present invention provide a real-time hybrid simulation test method and system for a fixed offshore wind turbine system, which can realize real-time and accurate simulation of the motion characteristics of a fixed offshore wind turbine system under the influence of natural conditions due to its response to natural conditions.

[0019] Figure 1 A flowchart of a real-time hybrid simulation test method for a stationary offshore wind turbine system according to an embodiment of the present invention is shown.

[0020] like Figure 1 As shown, the method of this embodiment may include operations S101 to S104.

[0021] In operation S101, the (i-1)th wind load parameter is reduced according to the wind parameter ratio, and the reduced (i-1)th wind load parameter is used to drive the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model. The offshore wind turbine system model is obtained by scaling down the actual offshore wind turbine system. The wind parameter ratio represents the proportional relationship between the wind load in the offshore wind turbine system model and the wind load in the actual offshore wind turbine system. i is a positive integer greater than 1.

[0022] In operation S102, the motion characteristics of the offshore wind turbine system model under the influence of the i-th simulated natural condition are collected. The offshore wind turbine system model includes saturated soil, a foundation structure set on the saturated soil, and a wind turbine structure set on the foundation structure. The i-th simulated natural condition includes the i-th wind load.

[0023] In operation S103, the i-th motion feature is expanded according to the characteristic proportional relationship to obtain the i-th expanded motion feature. The characteristic proportional relationship characterizes the proportional relationship between the motion features of the actual offshore wind turbine system and the motion features of the offshore wind turbine system model.

[0024] In operation S104, the target neural network is used to process the i-th amplified motion feature to obtain the i-th wind load parameter. The target neural network represents the nonlinear mapping relationship between the wind load parameters of the actual offshore wind turbine system and the corresponding motion features.

[0025] According to embodiments of the present invention, the offshore wind turbine system model can be a test substructure obtained by scaling down the actual offshore wind turbine system (hereinafter referred to as scaling). For example, based on the similarity criteria of the 1g gravity field dynamic model test, the actual offshore wind turbine system structure can be scaled down to obtain an offshore wind turbine system model including saturated soil, foundation structure, and wind turbine structure. The foundation structure may include cylindrical foundations and monopile foundations, etc., and the wind turbine structure may include towers and top concentrated mass, etc., with the top concentrated mass used to simulate the hub, blades, and nacelle assembly system of the actual wind turbine.

[0026] According to embodiments of the present invention, the i-th simulated natural condition may include wind load, wave load, and seismic force, etc. The i-th motion characteristic may be the motion characteristic of the offshore wind turbine system model under the influence of the i-th simulated natural condition, such as at least one of acceleration and displacement, etc. In some embodiments, the motion characteristics of the offshore wind turbine system model under simulated natural conditions can be collected by sensors.

[0027] According to an embodiment of the present invention, the i-th expanded motion characteristic can correspond to the motion characteristics of an actual offshore wind turbine system under natural conditions. The characteristic ratio can be the ratio between the motion characteristics of the actual offshore wind turbine system and the motion characteristics of the offshore wind turbine system model, determined according to the similarity criterion of a 1g (i.e., 1 gram) gravity field dynamic model test. Furthermore, in embodiments of the present invention, other ratios can also be determined according to the 1g gravity field dynamic model test similarity criterion, such as model ratios, wind parameter ratios, wave and current parameter ratios, and seismic parameter ratios, which can respectively refer to the ratio between the offshore wind turbine system model and the actual offshore wind turbine system, the ratio between the wind load of the offshore wind turbine system model and the wind load of the actual offshore wind turbine system, the ratio between the wave load of the offshore wind turbine system model and the wave load of the actual offshore wind turbine system, and the ratio between the seismic force of the offshore wind turbine system model and the seismic force of the actual offshore wind turbine system, etc.

[0028] For example, the scale ratio between the offshore wind turbine system model and the actual offshore wind turbine system can be used to construct experimental substructures. The scale ratio between the wind load on the offshore wind turbine system model and the actual offshore wind turbine system can be used to simulate the wind load applied to the offshore wind turbine system model based on the wind load of the actual offshore wind turbine system. Similarly, the scale ratio between the wave load on the offshore wind turbine system model and the actual offshore wind turbine system can be used to simulate the wave load applied to the offshore wind turbine system model based on the wave load of the actual offshore wind turbine system. Finally, the scale ratio between the seismic force on the offshore wind turbine system model and the actual offshore wind turbine system can be used to simulate the seismic force applied to the offshore wind turbine system model based on the seismic force of the actual offshore wind turbine system.

[0029] It should be noted that the first wind load parameter can be set based on the Kaimal wind spectrum. Subsequent wind load parameters following the first can be output by the aforementioned target neural network, and will not be elaborated upon here.

[0030] According to an embodiment of the present invention, the target neural network is a learning model constructed based on the natural parameters of the actual offshore wind turbine system and the motion features corresponding to the natural parameters. For example, it can be a deep neural network, such as a convolutional neural network (CNN), etc., to achieve the concept of the present invention, and the present invention does not limit it.

[0031] According to an embodiment of the present invention, the wind turbine structure in the above-described hybrid simulation test method may include a tower. The motion characteristics include the top acceleration characteristics of the tower, and the expanded motion characteristics include the expanded top acceleration characteristics.

[0032] According to an embodiment of the present invention, the wind load parameters for the next round can be determined based on the expanded motion characteristics, and then scaled down according to the wind parameter ratio. Using the motion characteristics of the offshore wind turbine system model under the influence of the current round of wind load, the top displacement characteristics of the offshore wind turbine system model for the current round are determined. After time-delay compensation of these top displacement characteristics, the compensated top displacement characteristics for the current round can be converted from target displacement to target force. The converted force (i.e., the top force characteristics of the current round) is used to compensate the scaled-down wind load parameters, obtaining the compensated wind load parameters for the next round. The scaled-down and compensated wind load parameters for the next round are provided to the complex boundary coordinated loading device to drive the complex boundary coordinated loading device to apply the next round of wind load to the offshore wind turbine system model.

[0033] According to embodiments of the present invention, a complex boundary coordinated loading device is a device for applying wind loads to an offshore wind turbine system model. For example, the applied wind load can be scaled down according to the similarity criteria of the 1g gravity field dynamic model test, and the scaled-down wind load parameters are provided to the complex boundary coordinated loading device. The complex boundary coordinated loading device can apply single / bidirectional wind loads, is lightweight and compact, and can also serve as the top concentrated mass of the offshore wind turbine system model (i.e., a point mass equivalent to the mass of components such as the hub, blades, and nacelle assembly system), simulating the hub, blades, and nacelle assembly system.

[0034] For example, a complex boundary coordinated loading device may include a frame, an electro-hydraulic servo valve, an electro-hydraulic servo actuator, a linear guide rail, and an inertial mass box. The frame, as the rigid support structure of the complex boundary coordinated loading device, bears the wind load applied by the electro-hydraulic servo actuator through columns and beams, ensuring structural stability during loading. The electro-hydraulic servo valve converts the controller's electrical signal into high-precision hydraulic flow and pressure, driving the electro-hydraulic servo actuator to achieve dynamic loading. The electro-hydraulic servo actuator converts the hydraulic flow and pressure output from the electro-hydraulic servo valve into mechanical thrust, directly applying the wind load to the offshore wind turbine system model. The linear guide rail, mounted on the frame's beams, provides high-precision linear guidance during the wind load loading process, enabling free horizontal displacement of the electro-hydraulic servo actuator. The inertial mass box simulates the inertial effects of rotating components in the offshore wind turbine system (such as the moment of inertia of blades / gearboxes), adjusting the mass distribution through counterweights to simulate the motion characteristics of an actual offshore wind turbine system under wind load.

[0035] According to an embodiment of the present invention, the motion characteristics of an offshore wind turbine system model set on saturated soil under wind load are collected, then amplified according to the characteristic ratio, and the amplified motion characteristics are input into a target neural network. Based on this, even if the foundation structure is set on saturated soil that has been damaged by seawater immersion, causing the offshore wind turbine system model to undergo complex motion under wind load due to the loose structure of the saturated soil, resulting in a non-linear relationship between wind load parameters and motion characteristics, the model parameters of the target neural network of the present invention can still converge during the fitting process. Therefore, the target neural network can still quickly and accurately output the corresponding wind load parameters based on the motion characteristics of this complex motion. Furthermore, in each round of simulation experiments, the wind load parameters output by the target neural network can be reduced, and the reduced wind load can be used to drive a complex boundary coordination loading device to apply accurate wind loads to the offshore wind turbine system model in a timely manner. This achieves real-time and accurate simulation of the motion characteristics of a fixed offshore wind turbine system under natural conditions. Furthermore, by using the reduced wind load parameters to drive the complex boundary coordination loading device to test the offshore wind turbine system model, which includes saturated soil, foundation structure, and wind turbine structure, the problem of scale inconsistency between the actual offshore wind turbine system and the offshore wind turbine system model is avoided, which is caused by the offshore wind turbine system model not accurately replicating the structure of the actual offshore wind turbine system.

[0036] Figure 2 A flowchart illustrating the training method of the target neural network in the real-time hybrid simulation test method for a fixed offshore wind turbine system according to an embodiment of the present invention is shown.

[0037] like Figure 2 As shown, the training method for the target neural network may include operations S21 to S23.

[0038] In operation S21, a simulation model of the offshore wind turbine system corresponding to the actual offshore wind turbine system is obtained. The model size of the offshore wind turbine system simulation model matches the actual size of the actual offshore wind turbine system.

[0039] During operation S22, the motion characteristics of the offshore wind turbine system simulation model are collected under natural conditions that match the natural condition parameters.

[0040] In operation S23, natural condition parameters are used as labels, and the motion characteristics of the offshore wind turbine system simulation model are used as training data. The initial neural network is trained using the labels and training data to obtain the target neural network.

[0041] According to an embodiment of the present invention, the offshore wind turbine system simulation model is a numerical substructure with actual dimensions (hereinafter referred to as full-scale) built on a wind turbine simulation platform using NREL FAST (Fatigue, Aerodynamics, Structures and Turbulence) software. The offshore wind turbine system simulation model may only include the wind turbine (i.e., hub, blades, and nacelle assembly system), excluding the tower and substructure, and is used to simulate the interaction between aerodynamics and the wind turbine blades.

[0042] According to embodiments of the present invention, natural condition parameters may include parameters such as wind spectrum model, wind speed, wave and current grade, wave and current velocity, and seismic acceleration time history. Natural conditions may include wind load, seismic motion, and wave load. The motion characteristics of the offshore wind turbine system simulation model may include the acceleration at the hub center in the offshore wind turbine system simulation model, and this acceleration can be used as the acceleration at the top of the tower. Thus, by inputting the above-mentioned natural condition parameters into the offshore wind turbine system simulation model, the acceleration at the top of the simulated tower can be obtained.

[0043] According to embodiments of the present invention, the training data may include the simulated tower top acceleration of an offshore wind turbine simulation model under natural conditions with matched natural condition parameters. For example, the training data can be input into an initial neural network to obtain the predicted natural condition parameters output by the initial neural network. Then, the loss value between the predicted natural condition parameters and the label can be calculated. Based on this loss value, the parameters of the initial neural network can be iteratively optimized using a backpropagation gradient optimization algorithm to obtain the trained target neural network. It should be noted that the trained target neural network only needs to meet the actual requirements, and this application does not impose any limitations on this.

[0044] According to an embodiment of the present invention, a full-scale simulation model of a fixed offshore wind turbine system is established using NREL FAST software. Parameters such as wind spectrum model, wind speed, turbulence level, wave and current velocity, and ground acceleration time history are input to obtain the motion characteristics of the acceleration at the top of the simulated tower. Then, the initial neural network is trained using these motion characteristics to obtain the model parameters of a convergent target neural network that characterizes the natural condition parameters of the actual offshore wind turbine system and the nonlinear mapping relationship between the motion characteristics corresponding to these natural condition parameters. This approach can at least partially avoid the problems of slow convergence, slow computation, and severe time delays in hybrid simulation experiments caused by the nonlinear behavior of the experimental substructure under natural conditions. It enables real-time interaction of time-domain data of motion characteristics between the offshore wind turbine system simulation model and the offshore wind turbine system model, achieving rapid and accurate calculation of the motion characteristics of the offshore wind turbine system model. For example, real-time interaction of time-domain data of motion characteristics between the offshore wind turbine system simulation model and the offshore wind turbine system model (at the top of the tower) can be achieved.

[0045] Figure 3 A flowchart illustrating the application of wind load according to an embodiment of the present invention is shown.

[0046] like Figure 3 As shown, the method for applying wind load in this embodiment may include operations S31 to S36.

[0047] In operation S31, the i-1th top acceleration feature after being expanded is processed by the target neural network to obtain the i-1th wind load parameter about the top of the tower.

[0048] In operation S32, based on the i-1 motion characteristics of the offshore wind turbine system model under the influence of the i-1 simulated natural conditions, the i-1 top displacement characteristics of the tower are determined.

[0049] In operation S33, time delay compensation is performed on the displacement feature of the (i-1)th top, and the force feature of the (i-1)th top is calculated based on the compensated displacement feature of the (i-1)th top.

[0050] In operation S34, the i-1th wind load parameter is reduced according to the wind parameter ratio, and the reduced i-1th wind load parameter is compensated based on the i-1th top stress characteristics to obtain the i-th compensated wind load parameter.

[0051] In operation S35, the i-th driving electrical signal is generated based on the i-th compensated wind load parameter.

[0052] In operation S36, the i-th driving electrical signal is used to drive the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model.

[0053] According to an embodiment of the present invention, the controller performs filtering and time-delay compensation on the collected top displacement features through commands such as band-stop filters and predicted displacements, converting the compensated top displacement features into corresponding top force features as time-delay compensation for the i-th wind load in operation S34. For example, time-delay compensation for the wind load parameters can be achieved based on the wind force corresponding to the top force features and wind load parameters. Then, the compensated wind load parameters can be converted into wind load parameters of the offshore wind turbine system model according to the characteristic proportional relationship of the wind load. The controller then converts the wind load parameters of the offshore wind turbine system model into drive electrical signals to drive the complex boundary coordinated loading device to apply wind load to the offshore wind turbine system model, thereby realizing the simulation of wind load. It should be noted that the above method for applying wind load can be applied to any round of simulation test, which will not be elaborated here.

[0054] According to an embodiment of the present invention, the motion characteristics of the offshore wind turbine system model under simulated natural conditions are enlarged according to a characteristic ratio (hereinafter referred to as scale expansion). The scaled-up motion characteristics of the offshore wind turbine system model are then input into a target neural network to calculate the scale expansion condition parameters, ensuring that the scale expansion condition parameters of the offshore wind turbine system model are at the full scale. This improves the response consistency between the offshore wind turbine system model and the actual offshore wind turbine system without changing the Reynolds number. Through a complex boundary coordination loading device, an iterative calculation and data closed loop is formed between the offshore wind turbine system simulation model and the actual offshore wind turbine system model.

[0055] According to an embodiment of the present invention, applying the i-th wind load to an offshore wind turbine system model by driving a complex boundary coordinated loading device may include: driving the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model in at least one horizontal direction, either a first horizontal direction or a second horizontal direction; the first horizontal direction intersects (e.g., perpendicularly) the second horizontal direction. The first horizontal direction and the second horizontal direction can be any two different horizontal directions. Based on this, the complex boundary coordinated loading device can achieve dynamic application of single / bidirectional wind loads, enabling interaction between the target neural network and the response data of the offshore wind turbine system model.

[0056] Thus, according to an embodiment of the present invention, the complex boundary coordinated loading device can realize the dynamic application of single / bidirectional wind loads and realize the interaction between the target neural network and the offshore wind turbine system model on the motion characteristics of the top of the wind turbine structure, thereby alleviating the problem of convergence difficulty in some schemes of vibration table hybrid simulation test.

[0057] According to an embodiment of the present invention, saturated soil is immersed in a simulated liquid, such as water, specifically seawater; the i-th simulated natural condition further includes an i-th wave current load for the simulated liquid and an i-th seismic motion for the saturated soil. The hybrid simulation test method further includes at least one of the following: based on the i-th wave current load parameter, driving a wave current generating device to generate the i-th wave current load; the i-th wave current load parameter is obtained by reducing the wave current load of the actual offshore wind turbine system according to the wave current parameter ratio; the wave current parameter ratio characterizes the proportional relationship between the wave current load of the offshore wind turbine system model and the wave current load of the actual offshore wind turbine system; based on the i-th seismic parameter, driving an underwater shaking table to generate the i-th seismic motion; the i-th seismic parameter is obtained by reducing the seismic force on the actual offshore wind turbine system according to the seismic parameter ratio; the seismic parameter ratio characterizes the proportional relationship between the seismic parameters of the offshore wind turbine system model and the seismic parameters of the actual offshore wind turbine system.

[0058] According to an embodiment of the present invention, saturated soil soaking in a simulated liquid can simulate the actual natural conditions of an offshore wind turbine system, and can reduce the impact of model structure stiffness degradation and soil plastic deformation on the simulation results in some schemes.

[0059] According to an embodiment of the present invention, the wave and current load is applied by the wave and current generating device, and the ground motion is applied by the underwater shaking table. The wave and current load and the ground motion can be obtained by scaling down the load of the actual offshore wind turbine system according to the characteristic proportional relationship based on the similarity criterion of the 1g gravity field dynamic model test.

[0060] According to embodiments of the present invention, this method can improve the accuracy of simulating the motion characteristics of offshore wind turbine system models under the coupled effects of wind, wave, current, and earthquake loads. Offshore wind turbine systems must simultaneously withstand the dynamic coupling effects of multiple loads, including wind, wave, current, and earthquake. Single-load analysis may underestimate the structural response under extreme conditions. Joint simulation can effectively capture the nonlinear superposition effect between loads. Thus, applying the method of the present invention can improve the accuracy of real-time hybrid simulation of stationary offshore wind turbine systems.

[0061] According to an embodiment of the present invention, a real-time hybrid simulation test system for a fixed offshore wind turbine system includes an offshore wind turbine system model, a complex boundary coordination loading device, and a controller. The offshore wind turbine system model includes saturated soil, a foundation structure situated on the saturated soil, and a wind turbine structure situated on the foundation structure. The offshore wind turbine system model is obtained by scaling down an actual offshore wind turbine system. The controller is used to execute the aforementioned hybrid simulation test method.

[0062] According to an embodiment of the present invention, the controller can convert the wind load parameters of the offshore wind turbine system model into electrical signals and drive the complex boundary coordinated loading device to load the offshore wind turbine system model, thereby collecting the tower top acceleration generated by the offshore wind turbine system model; the controller can also return the tower top acceleration generated by the offshore wind turbine system model to the numerical substructure (i.e., input target neural network), and after scaling the motion features according to the feature ratio, the offshore wind turbine system simulation model calculates the (i+1)th wind load based on the motion features.

[0063] According to an embodiment of the present invention, the above-described hybrid simulation test system further includes a wave-current generating device and an underwater vibration table.

[0064] According to an embodiment of the present invention, the wave-current generating device can apply wave-current loads, the underwater shaking table can apply seismic motion, and the wind load applied by the complex boundary coordination loading device can be used to simulate the influence of wind-wave-current-earthquake multi-load coupling on the motion characteristics of the offshore wind turbine system model.

[0065] Figure 4 A schematic diagram of the operation of a real-time hybrid simulation test system for a stationary offshore wind turbine system according to an embodiment of the present invention is shown.

[0066] According to an embodiment of the present invention, the real-time hybrid simulation test method for a fixed offshore wind turbine system includes a simulation model of the offshore wind turbine system (hereinafter referred to as the numerical substructure) and a model of the offshore wind turbine system (hereinafter referred to as the experimental substructure). The model size of the numerical substructure matches the actual size of the actual offshore wind turbine system, while the model size of the experimental substructure is obtained by scaling down the actual size of the actual offshore wind turbine system according to a characteristic proportional relationship.

[0067] According to an embodiment of the present invention, a full-scale numerical substructure is established using NREL FAST software. Parameters such as wind spectrum model, wind speed, turbulence level, wave and current velocity, and seismic acceleration time history are input to obtain the motion characteristics of the acceleration at the top of the simulated tower. Based on a target neural network, a nonlinear mapping relationship is established between natural condition parameters and motion characteristics (i.e., acceleration at the top of the tower), enabling rapid and accurate calculation of the motion characteristics of the numerical substructure. Wave and current loads are applied by a wave and current generation device, seismic input is applied by an underwater shaking table, and wind loads are applied by a complex boundary coordinated loading device.

[0068] According to an embodiment of the present invention, the pulsating wind load of the actual offshore wind turbine system (i.e., the wind load mentioned above) is scaled down according to a characteristic ratio to obtain the wind load of the test substructure. The controller then converts this into a driving electrical signal to drive the boundary coordination loading device (i.e., the complex boundary coordination loading device mentioned above) to load the test substructure, and collects the tower top acceleration generated by the test substructure. The tower top acceleration generated by the test substructure is returned to the numerical substructure through the controller. After scaling up the tower top acceleration according to the characteristic ratio, the numerical substructure calculates the wind load for the next round based on the scaled-up tower top acceleration, thus ensuring that all numerical substructures are full-scale, without changing the Reynolds number, and guaranteeing the consistency between the simulation results and the prototype response. The controller uses commands such as band-stop filters and predicted displacement to filter and compensate for the time delay of the collected top displacement features, converting the compensated top displacement features into corresponding top force features as time delay compensation for the next round of wind load. Then, the expanded wind load parameters can be converted into wind load parameters of the offshore wind turbine system model according to the characteristic proportional relationship of wind load. The controller then converts the wind load parameters of the offshore wind turbine system model into driving electrical signals to drive the boundary coordination loading device to apply wind load to the offshore wind turbine system model, thereby realizing the simulation of wind load and finally forming an iterative calculation and data closed loop between the numerical substructure and the experimental substructure.

[0069] According to embodiments of the present invention, the above-described hybrid simulation test method can alleviate the Froude-Reynolds scaling inconsistency problem existing in some shaking table tests, and improve the accuracy of simulating the motion characteristics of offshore wind turbines under the combined effects of wind, wave, current, and earthquake. The use of a complex boundary coordination loading device enables accurate dynamic application of single / bidirectional wind loads, and allows for real-time interaction of acceleration time-domain data between the numerical substructure and the experimental substructure at the top of the tower, alleviating the convergence difficulties in some hybrid shaking table simulation tests. The introduction of a target neural network alleviates the problems of slow convergence, slow computation, and severe time delays in hybrid simulation tests caused by structural nonlinear behavior, enabling real-time interaction between the numerical substructure and the experimental substructure.

[0070] Those skilled in the art will understand that the features described in the various embodiments of the present invention can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

[0071] The embodiments of the present invention have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of the invention. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the invention, and all such substitutions and modifications should fall within the scope of the invention.

Claims

1. A real-time hybrid simulation test method for a fixed offshore wind turbine system, characterized in that, include: According to the wind parameter ratio, the (i-1)th wind load parameter is reduced in size, and the reduced (i-1)th wind load parameter is used to drive the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model; the offshore wind turbine system model is obtained by reducing the size of the actual offshore wind turbine system; the wind parameter ratio represents the proportional relationship between the wind load of the offshore wind turbine system model and the wind load of the actual offshore wind turbine system; i is a positive integer greater than 1; The i-th motion characteristics of the offshore wind turbine system model under the influence of the i-th simulated natural conditions are collected. The offshore wind turbine system model includes saturated soil, a foundation structure set on the saturated soil, and a wind turbine structure set on the foundation structure; the i-th simulated natural condition includes the i-th wind load; According to the characteristic ratio relationship, the i-th motion feature is enlarged to obtain the i-th enlarged motion feature; The characteristic proportional relationship characterizes the proportional relationship between the motion characteristics of the actual offshore wind turbine system and the motion characteristics of the offshore wind turbine system model; The i-th amplified motion feature is processed using a target neural network to obtain the i-th wind load parameter; the target neural network characterizes the nonlinear mapping relationship between the wind load parameter of the actual offshore wind turbine system and the motion feature corresponding to the wind load parameter. The wind turbine structure includes a tower; the motion characteristics include the top acceleration characteristics of the tower; the expanded motion characteristics include the expanded top acceleration characteristics. According to the wind parameter ratio, the (i-1)th wind load parameter is reduced, and the reduced (i-1)th wind load parameter is used to drive the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model, including: Based on the motion characteristics of the offshore wind turbine system model under the influence of the (i-1)th simulated natural conditions, the (i-1)th top displacement characteristics of the tower are determined. Time delay compensation is performed on the (i-1)th top displacement feature, and the (i-1)th top force feature is calculated based on the compensated (i-1)th top displacement feature; According to the wind parameter ratio, the i-1th wind load parameter is reduced, and the reduced i-1th wind load parameter is compensated based on the i-1th top stress characteristics to obtain the i-th compensated wind load parameter. Based on the i-th compensated wind load parameters, generate the i-th driving electrical signal; The i-th driving electrical signal is used to drive the complex boundary coordinated loading device to apply the i-th wind load to the offshore wind turbine system model.

2. The method according to claim 1, characterized in that, The i-th amplified motion feature is processed using a target neural network to obtain the i-th wind load parameter, including: By processing the amplified top acceleration features using the target neural network, the i-th wind load parameter for the top of the tower is obtained.

3. The method according to claim 1, characterized in that, Driving the complex boundary coordination loading device to apply the i-th wind load to the offshore wind turbine system model includes: The complex boundary coordination loading device is driven to apply the i-th wind load to the offshore wind turbine system model in at least one of the first or second horizontal directions; the first horizontal direction intersects with the second horizontal direction.

4. The method according to any one of claims 1 to 3, characterized in that, The saturated soil is immersed in a simulated liquid; the i-th simulated natural condition also includes an i-th wave load on the simulated liquid and an i-th ground motion on the saturated soil; The method further includes at least one of the following: Based on the i-th wave current load parameter, the wave current generating device is driven to generate the i-th wave current load; the i-th wave current load parameter is obtained by reducing the wave current load of the actual offshore wind turbine system according to the wave current parameter ratio; the wave current parameter ratio characterizes the proportional relationship between the wave current load of the offshore wind turbine system model and the wave current load of the actual offshore wind turbine system. Based on the i-th earthquake parameter, an underwater shaking table is driven to generate the i-th earthquake motion; the i-th earthquake parameter is obtained by reducing the seismic force experienced by the actual offshore wind turbine system according to the earthquake parameter ratio; The seismic parameter ratio characterizes the proportional relationship between the seismic parameters of the offshore wind turbine system model and the seismic parameters of the actual offshore wind turbine system.

5. The method according to any one of claims 1 to 3, characterized in that, The target neural network is trained in the following manner: Obtain a simulation model of the offshore wind turbine system corresponding to the actual offshore wind turbine system; the model size of the offshore wind turbine system simulation model matches the actual size of the actual offshore wind turbine system; The motion characteristics of the offshore wind turbine system simulation model are collected under natural conditions that match the natural condition parameters. Using the natural condition parameters as labels and the motion characteristics of the offshore wind turbine system simulation model as training data, the initial neural network is trained using the labels and the training data to obtain the target neural network.

6. The method according to claim 5, characterized in that, The offshore wind turbine system simulation model was built using a wind turbine simulation platform.

7. A real-time hybrid simulation test system for a fixed offshore wind turbine system, characterized in that, include: The offshore wind turbine system model includes saturated soil, a foundation structure set on the saturated soil, and a wind turbine structure set on the foundation structure; the offshore wind turbine system model is obtained by scaling down the actual offshore wind turbine system. Complex boundary coordination loading device; as well as A controller for performing a real-time hybrid simulation test method for a stationary offshore wind turbine system as described in any one of claims 1 to 6.

8. The system according to claim 7, characterized in that, Also includes: A wave flow generating device, used to generate wave flow loads under the control of a controller; An underwater vibration table is used to generate ground vibrations under the control of a controller.

Images