System and method for controlling a wind turbine

Abstract

A system and method for controlling a wind turbine is provided. In one embodiment, a wind classification module of the controller determines a current aerodynamic state of the wind resource based at least in part on a current operating dataset of the wind turbine. The current operating dataset is indicative of a current operation of the wind turbine. A configuration intelligence module of the controller then generates an estimated configuration for the turbine estimator module and a predictive control configuration for the predictive control module based at least in part on the current aerodynamic state. The operation of the wind turbine is emulated via the turbine estimator module to generate an initial control state for the predictive control module. The predictive control module determines a predicted performance of the wind turbine over a prediction horizon based on the initial control state and the predictive control configuration. The predictive control module generates a set point for at least one actuator of the wind turbine based on the predicted performance, and the operating state of the wind turbine is influenced via the at least one actuator according to the set point.

Description

[Technical Field]

[0001] FIELD OF THE DISCLOSURE The present disclosure relates generally to power generation, and more particularly to systems and methods for controlling wind turbines via predictive control. [Background technology]

[0002] Wind power is generally considered one of the cleanest and most environmentally friendly energy sources currently available. As such, wind turbines have become a popular choice. Modern wind turbines typically include a tower, a generator, a gearbox, a nacelle, and one or more rotor blades. The nacelle contains a rotor assembly coupled to a gearbox and a generator. The rotor assembly and gearbox are mounted on a bedplate support frame installed within the nacelle. The rotor blades capture the kinetic energy of the wind using known airfoil principles. The rotor blades transfer the kinetic energy in the form of rotational energy to rotate a shaft connecting the rotor blades to the gearbox, or directly to a generator if a gearbox is not used. The generator then converts the mechanical energy into electrical energy, which is transmitted to a converter and / or transformer housed in the tower before being deployed to the utility grid. Modern wind power generation systems typically take the form of a wind farm having a number of wind turbine generators operable to supply power to a transmission system that supplies power to a power grid.

[0003] Current wind turbine controls typically use generator torque and blade pitch actuators to maintain mechanical loads below design limits through a wide range of wind conditions and / or fault scenarios. During normal operation, the most basic control function is to regulate power output and rotor speed according to predetermined schedules. These schedules are generally dependent on the wind speed affecting the wind turbine. At low wind speeds, rotor speed is regulated by torque, orienting the blades to capture the most wind power. At higher wind speeds, rotor speed can be regulated to its rated value by pitching the blades away from the wind. Additionally, power can be regulated by generator torque.

[0004] In addition to the most basic control functions, current wind turbine control may include multiple functions, or control loops, to mitigate loads due to sudden wind gusts, increased wind shear, and / or faults in the system. These protection loops / schemes may consist of out-of-phase cyclic pitch, rapid pitching (e.g., pitching away from the wind to limit loads), and / or alternating pitching in and out of the wind. The real-time data required to regulate control loops / schemes and / or trigger protection loops can be obtained using estimation techniques (e.g., Kalman filters) based on simplified turbine models, for example. The complexity of the simplified turbine models may be limited by the computational limitations of known industrial control platforms.

[0005] However, using simplified models to acquire data can limit the controller's ability to operate the wind turbine in an optimal manner. For example, simplified models commonly employed in turbine estimation / control are simplistic and may not accurately capture the wind conditions affecting the structure and / or wind turbine. This can result in conservative, suboptimal turbine control. As an additional example, many load-reduction schemes / loops are designed independently of each other, with each loop / scheme sending a pitch command. As a result, the pitch system may experience unnecessary wear and be susceptible to instabilities caused by rate saturations. Furthermore, an increasing number of protection loops / schemes can result in complex dynamic interactions, making controller tuning difficult and resulting in suboptimal performance. Furthermore, protection loops / schemes may indirectly handle turbine constraints by choosing surrogate approximations, which risks significant constraint violations, or may operate conservatively, leading to suboptimal performance. To give a further example, commonly adopted simplified models may not be suitable for modeling the increased dynamic coupling due to the more flexible structure of modern wind turbines.

[0006] Accordingly, there is a continuing need in the art for new and improved systems and methods that address the aforementioned problems. Thus, the present disclosure is directed to a system and method for controlling a wind turbine. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] US Patent Application Publication No. 2017 / 350369 Summary of the Invention

[0008] Aspects and advantages of the present invention will be set forth in part in the description that follows, or will be obvious from the description, or may be learned through practice of the invention.

[0009] In one aspect, the disclosure is directed to a method for controlling a wind turbine. The method may include determining, via a wind classification module of a controller, a current aerodynamic state of a wind resource based at least in part on a wind turbine current operational dataset indicative of a current operation of the wind turbine. A configuration intelligence module of the controller may be utilized to generate an estimator configuration for a turbine estimator module based at least in part on the current aerodynamic state. The turbine estimator module of the controller may emulate operation of the wind turbine to generate a control initial state for a predictive control module. The control initial state may include a modeled current operating state of a plurality of components of the wind turbine. The method may also include generating, via the configuration intelligence module, a predictive control configuration for the predictive control module based at least in part on the current aerodynamic state. The predictive control module may be employed to determine a predicted performance of the wind turbine over a predictive interval based on the control initial state and the predictive control configuration. The predictive control module may also be employed to generate a setpoint for at least one actuator of the wind turbine based on the predicted performance. Further, the method may include affecting an operating condition of the wind turbine via the actuator according to the setpoint.

[0010] These and other features, aspects, and advantages of the present invention will become better understood with reference to the following description and appended claims. The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention. [Brief explanation of the drawings]

[0011] A full and enabling disclosure of the present invention, including the best mode thereof, directed to one of ordinary skill in the art, is set forth in the following specification, which makes reference to the accompanying drawings, in which: [Figure 1] 1 is a perspective view of an embodiment of a wind turbine. [Figure 2] FIG. 2 is a perspective interior view of an embodiment of a nacelle of the wind turbine of FIG. 1. [Figure 3] FIG. 1 illustrates a block diagram of an embodiment of a controller for use with a wind turbine. [Figure 4] 1 is a schematic diagram of a system for controlling a wind turbine; [Figure 5] FIG. 1 is a schematic diagram of a control logic for controlling a wind turbine. [Figure 6] FIG. 6 is a schematic diagram of a portion of the control logic of FIG. 5 for controlling the wind turbine. [Figure 7] 6 shows a schematic diagram of a portion of the control logic of FIG. 5 for controlling a wind turbine in the presence of nominal wind conditions. [Figure 8] 6 is a schematic diagram of a portion of the control logic of FIG. 5 for controlling the wind turbine in the presence of a first wind condition. [Figure 9] FIG. 6 is a schematic diagram of a portion of the control logic of FIG. 5 for controlling the wind turbine.

[0012] Repeat use of reference characters in the present specification and drawings is intended to represent same or analogous features or elements of the invention. DETAILED DESCRIPTION OF THE INVENTION

[0013] Reference will now be made in detail to the embodiments of the invention, one or more examples of which are illustrated in the drawings. Each example is provided by way of explanation of the invention, not limitation of the invention. Indeed, it will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the scope or spirit of the invention. For example, features illustrated or described as part of one embodiment can be used with another embodiment to yield a still further embodiment. Thus, it is intended that the present invention cover such modifications and variations as come within the scope of the appended claims and their equivalents.

[0014] As used herein, the terms "first," "second," and "third" may be used interchangeably to distinguish one component from another and are not intended to denote the position or importance of the individual components.

[0015] Terms such as "coupled," "fixed," "attached to," and the like, unless otherwise specified herein, refer to both direct coupling, fixing, or attachment, and indirect coupling, fixing, or attachment through one or more intermediate components or features.

[0016] Approximate terms used throughout this specification and claims are intended to modify any quantitative expression that can be permissibly varied without resulting in a change in the basic function to which it pertains. Thus, values ​​modified by terms such as "about," "approximately," and "substantially" are not limited to the exact value specified. In at least some instances, the approximate term may correspond to the precision of an instrument for measuring a value or the precision of a method or machine for constructing or manufacturing a component and / or system. For example, the approximate term may refer to being within a 10% margin.

[0017] Throughout this specification and claims, range limitations are combinable and interchangeable, and such ranges include all subranges specified and contained therein, unless the context or language dictates otherwise. For example, all ranges disclosed herein are inclusive of the endpoints, and the endpoints are independently combinable with each other.

[0018] In general, the present disclosure is directed to systems and methods for controlling wind turbines. In particular, the systems and methods disclosed herein can leverage the accuracy and numerical efficiency of online aeroelastic models (e.g., aeroelastic models executed in real time during wind turbine operation) to provide detailed physical information about the wind turbine. This detailed physical information can be used online for estimation, control, and / or fault management. This facilitates full utilization of aeroelastic predictive control to provide baseline core control functionality. Furthermore, the present systems and methods can employ multiple physics-based aeroelastic estimators in real time to determine aerodynamic conditions affecting the wind turbine and / or wind turbine operating conditions (e.g., fault conditions). This determination can be further facilitated by the use of machine learning algorithms. The use of machine learning algorithms can be facilitated by the integration of advanced computing hardware and software (e.g., parallel computing that can be performed via multi-core architectures) in industrial control frameworks.

[0019] Advantages of the present systems and methods may manifest in utilizing a number of aeroelastic estimators to determine a current aerodynamic state and / or a current operating state of the wind turbine based on an operational dataset (e.g., without environmental data), tuning a predictive control module, and generating at least one setpoint based on the predicted performance of the wind turbine. For example, the configuration intelligence module may select gain tuning based at least in part on the current aerodynamic state and generate an estimator configuration for the turbine estimator module and a predictive control configuration for the predictive control module. The turbine estimator module may emulate operation of the wind turbine to generate initial control states for the predictive control module. The predictive control module may then determine predicted performance of the wind turbine over a prediction horizon based on the initial control state and the predictive control configuration. The predictive control module may generate a setpoint for at least one actuator of the wind turbine based on the predicted performance and affect the operating state of the wind turbine via the actuator according to the setpoint.

[0020] In other words, the effect of wind on a wind turbine can be reflected in the wind turbine's operational response. Therefore, an aeroelastic model tuned to various estimated wind conditions (e.g., nominal, turbulent, shear, veer, gust, etc.) can be employed to determine the wind aerodynamic conditions that result in the monitored operational response without the need to directly monitor the aerodynamic conditions. The aerodynamic conditions, including the current wind conditions (e.g., nominal, turbulent, shear, gust, etc.) determined from the operational dataset can then be utilized to select appropriate gain tuning configured to facilitate an optimal response of the wind turbine in light of the determined current wind conditions. The gain tuning can be reflected in the estimator configuration and the predictive control configuration. When tuned to the current wind conditions, the turbine estimator module can emulate the current response of the wind turbine to the current wind conditions. This emulation can serve as a starting condition (e.g., control initial state) for the predictive control module. Similarly adjusted for current wind conditions, the predictive control module can predict how wind turbine operation will vary from starting conditions over a future prediction horizon. Based on this future performance prediction, the predictive control module generates setpoints for the wind turbine actuators that may maintain the current operating state of the wind turbine or may seek to influence the operating state to optimize wind turbine performance.

[0021] Referring now to the drawings, FIG. 1 illustrates a perspective view of one embodiment of a wind turbine 100 according to the present disclosure. The wind turbine 100 may generally include a tower 102 extending from a support surface 104 (e.g., the ground), a nacelle 106 mounted to the tower 102, and a rotor 108 coupled to the nacelle 106. The rotor 108 may include a rotatable hub 110 and at least one rotor blade 112 coupled to and extending outward from the hub 110. For example, in the illustrated embodiment, the rotor 108 includes three rotor blades 112. However, in additional embodiments, the rotor 108 may include more or fewer than three rotor blades 112. Each rotor blade 112 may be spaced apart relative to the hub 110 to enable the rotor 108 to rotate and convert kinetic energy from the wind into usable mechanical energy, and thus electrical energy. For example, the hub 110 may be rotatably coupled to a generator 118 (FIG. 2) disposed within the nacelle 106 to enable the generation of electrical energy.

[0022] Wind turbine 100 may also include a controller 200. In one embodiment, controller 200 may be configured as a turbine controller 202 centrally located within nacelle 106. However, in other embodiments, controller 200 may be located within other components of wind turbine 100 or at a location external to the wind turbine. Additionally, turbine controller 202 may be communicatively coupled to any number of components of wind turbine 100 for controlling the components and / or a farm controller configured to control multiple wind turbines 100. As such, controller 200 may include a computer or other suitable processing device.

[0023] In one embodiment, controller 200 may include suitable computer-readable instructions that, when executed, configure turbine controller 202 to perform a variety of different functions, such as receiving, transmitting, and / or executing wind turbine control / command signals. Additionally, wind turbine 100 may include a number of actuators 160 ( FIG. 2 ) configured to execute various command signals (e.g., setpoints) and affect the operating state of wind turbine 100. It should be understood that, as used herein, “operating state” may refer to the physical configuration, orientation, and / or operating status of wind turbine 100 or its components.

[0024] Referring now to FIG. 2 , a simplified internal view of one embodiment of the nacelle 106 of the wind turbine 100 shown in FIG. 1 is illustrated. As shown, a generator 118 may be coupled to the rotor 108 to generate electrical power from the rotational energy generated by the rotor 108. For example, as shown in the illustrated embodiment, the rotor 108 may include a rotor shaft 122 coupled for rotation to a hub 110. The rotor shaft 122 may be rotatably supported by main bearings 144. The rotor shaft 122 may, in turn, be rotatably coupled to a high-speed shaft 124 of the generator 118 via a gearbox 126 coupled to a bedplate support frame 136. As generally understood, the rotor shaft 122 may provide a low-speed, high-torque input to the gearbox 126 in response to rotation of the rotor blades 112 and hub 110 during operation of the wind turbine 100. Gearbox 126 may then be configured to convert the low-speed, high-torque input to a high-speed, low-torque output to drive high-speed shaft 124 and thus generator 118 during operation of wind turbine 100.

[0025] Each rotor blade 112 also includes a pitch control mechanism 120 configured to rotate each rotor blade 112 about its pitch axis 116. Each pitch control mechanism 120 may include a pitch drive motor 128 (which may be considered an actuator 160), a pitch drive gearbox 130, and a pitch drive pinion 132. In such an embodiment, pitch drive motor 128 may be coupled to pitch drive gearbox 130 such that pitch drive motor 128 provides a mechanical force to pitch drive gearbox 130. Similarly, pitch drive gearbox 130 may be coupled to pitch drive pinion 132 for rotation thereof. Pitch drive pinion 132 may, in turn, be in rotational engagement with pitch bearing 134 coupled between hub 110 and the corresponding rotor 112 such that rotation of pitch drive pinion 132 causes rotation of pitch bearing 134. Thus, in such an embodiment, rotation of pitch drive motor 128, such as in response to a set point, drives pitch drive gearbox 130 and pitch drive pinion 132, thereby rotating pitch bearing 134 and one or more rotor blades 112 about pitch axis 116.

[0026] It should be appreciated that pitching one or more rotor blades 112 about the pitch axis 116 may change the angle of attack between the one or more rotor blades 112 and the apparent wind. Accordingly, the one or more rotor blades 112 may pitch to feather when the one or more rotor blades 112 rotate about the pitch axis 116 to align with the apparent wind, and may pitch to power when the one or more rotor blades 112 rotate toward a direction generally perpendicular to the apparent wind. Furthermore, it should be appreciated that pitching to feather typically depowers the one or more rotor blades 112 as a result of the resulting reduced lift. Therefore, pitching the rotor blades 112 about the pitch axis 116 according to a set point may affect the operating conditions of the wind turbine 100.

[0027] Similarly, wind turbine 100 may include one or more yaw drive mechanisms 138 communicatively coupled to controller 200, with each yaw drive mechanism 138 configured to change the angle of nacelle 106 relative to the wind (e.g., by engaging a yaw bearing 140 of wind turbine 100). It should be appreciated that controller 200 may direct the yaw of nacelle 106 and / or the pitching of rotor blades 112 to aerodynamically orient wind turbine 100 relative to the wind acting on wind turbine 100, thereby facilitating power generation.

[0028] 2 , wind turbine 100 may include at least one sensor system 154 having one or more operational sensors 158 capable of monitoring the operation of wind turbine 100. Operational sensors 158 may be configured to detect performance of wind turbine 100 in response to environmental conditions, for example. For example, operational sensors 158 may be rotational speed sensors, position sensors, acceleration sensors, and / or output sensors operably coupled to controller 200.

[0029] In an embodiment, one or more motion sensors 158 may be directed at or integral with any suitable component of wind turbine 100. For example, motion sensor 158 may be directed at rotor shaft 122 and / or generator 118 of wind turbine 100. One or more motion sensors 158 may collect data indicative of the rotational speed and / or rotational position of rotor shaft 122, or any other shaft of wind turbine 100, and thus rotor 108, or a pump, in the form of a rotor speed, a rotor azimuth, and / or any other suitable measurement. The one or more motion sensors 158 may, in embodiments, be an analog tachometer, a DC tachometer, an AC tachometer, a digital tachometer, a contact tachometer, a non-contact tachometer, or a time and frequency tachometer. In one embodiment, the motion sensor 158 may be an encoder, such as, for example, an optical encoder.

[0030] In embodiments, the one or more operational sensors 158 of the one or more sensor systems 154 may be an ammeter, a voltmeter, an ohmmeter, and / or any other suitable sensor for monitoring the electrical condition of the wind turbine 100. Additionally, in embodiments, the one or more operational sensors 158 may be a strain gauge, a proximity sensor, and / or any other suitable sensor configured to detect displacement of the wind turbine 100 or its components.

[0031] In additional embodiments, one or more motion sensors 158 of one or more sensor systems 154 may be a pitch sensor. In this manner, controller 200 may receive a pitch indication for one or more rotor blades 112 of wind turbine 100 via one or more motion sensors 158 operatively coupled to pitch control mechanism 120. Controller 200 may consider the pitch setpoint indication in light of the operation of wind turbine 100 to determine whether the operating condition is such that the pitch of rotor 108 may be changed to meet a desired power generation output.

[0032] It should also be understood that, as used herein, the term "monitor" and variations thereof indicate that various sensors in wind turbine 100 may be configured to provide direct measurements of monitored parameters or indirect measurements of such parameters. Thus, the sensors described herein may be used, for example, to generate signals related to the monitored parameters, which may then be utilized by controller 200 to determine the status or response of wind turbine 100 and / or its components.

[0033] 3-9, multiple embodiments of a system 300 for controlling a wind turbine 100 according to the present disclosure are presented. As particularly shown in FIG. 3, a schematic diagram of one embodiment of suitable components that may be included within the system 300 is shown. For example, as shown, the system 300 may include a controller 200. The controller 200 may be configured to determine an aerodynamic state of a wind resource from operational data, tune a turbine estimator and a predictive control module based on the aerodynamic state, and determine setpoints based on predicted performance of the wind turbine. Furthermore, the controller 200 may be configured to model predicted operational responses of the wind turbine to a plurality of potential aerodynamic states of a wind resource that affect the wind turbine. As such, the controller 200 may be employed offline and / or in real time (e.g., while the turbine is operating). Thus, in one embodiment, controller 200 may be a single component located with wind turbine 100. However, in further embodiments, controller 200 may encompass multiple components located with wind turbine 100. In further embodiments, controller 200 may include additional components located remotely from wind turbine 100.

[0034] The controller 200 and / or turbine controller 202 may be communicatively coupled to one or more sensor systems 154, and thus one or more operational sensors 158. Further, as shown, the controller 200 may include one or more processors 206 and associated memory devices 208 configured to perform various computer-implemented functions (e.g., executing methods, steps, calculations, etc., as disclosed herein and storing associated data). Additionally, the controller 200 may also include a communications module 210 to facilitate communications between the controller 200 and various components of the wind turbine 100. Furthermore, the communications module 210 may include a sensor interface 212 (e.g., one or more analog-to-digital converters) that enable signals transmitted from the sensors 158 to be converted into signals that can be understood and processed by the processor 206. It should be understood that the one or more sensors 158 may be communicatively coupled to the communications module 210 using any suitable means. For example, the sensors 158 may be coupled to the sensor interface 212 via a wired connection. However, in other embodiments, one or more sensors 158 may be coupled to sensor interface 212 via a wireless connection, such as using any suitable wireless communication protocol known in the art. Additionally, communication module 210 may also be operably coupled to at least one actuator 160 configured to implement a control action directed by a command signal (e.g., a control vector or setpoint).

[0035] As used herein, the term "processor" refers not only to integrated circuits referred to in the art as being included in a computer, but also to controllers, microcontrollers, microcomputers, programmable logic controllers (PLCs), application-specific integrated circuits, and other programmable circuits. Furthermore, the one or more memory devices 208 may generally be comprised of one or more memory elements, including, but not limited to, computer-readable media (e.g., random access memory (RAM)), computer-readable non-volatile media (e.g., flash memory), floppy disks™, compact disk read-only memories (CD-ROMs), magneto-optical disks (MODs), digital versatile disks (DVDs), and / or other suitable memory elements. Such one or more memory devices 208 may generally be configured to store suitable computer-readable instructions that, when executed by the one or more processors 206, configure the controller 200 / turbine controller 202 to perform various functions in accordance with the methods disclosed herein, as well as various other suitable computer-implemented functions.

[0036] 3-9 illustrate various aspects of a system 300 for controlling wind turbine 100. Accordingly, in one embodiment, controller 200 may be configured to implement a process 400 for determining a current aerodynamic state 302 of a wind resource affecting wind turbine 100. Current aerodynamic state 302 may be determined, for example, via wind classification module 214 of controller 200. Current aerodynamic state 302 may be based at least in part on a current operating dataset 304. Current operating dataset 304 may indicate a current operation 305 of wind turbine 100 in response to the wind resource (e.g., real-time operation updated every 40 milliseconds). The controller 200 may also utilize a configuration intelligence module 216 to determine an estimator configuration 306 for a turbine estimator module 218 of the controller 200 and a predictive control configuration 308 for a predictive control module 220 of the controller 200 based at least in part on the current aerodynamic state 302. As depicted at 309, the turbine estimator module 218 may emulate the operation of the wind turbine 100 to generate a control initial state 310 for the predictive control module 220. The control initial state 310 may include a modeled current operating state of multiple components of the wind turbine 100. The predictive control module 220 may then utilize at least the control initial state 310 and the predictive control configuration 308 to determine a predicted performance 312 of the wind turbine 100 over a prediction horizon 314. The controller 200 may further utilize a predictive control module 220 to generate at least one setpoint 316 for at least one actuator 160 of the wind turbine 100 based on the predicted performance 312 .Additionally, the system 300 can affect an operating state 318 of the wind turbine 100 via the actuators 160 according to a setpoint 316 .

[0037] In one embodiment, wind classification module 214 may include a first plurality of aeroelastic estimators 320. As used herein, the term “aeroelastic estimator” may refer to an estimator that utilizes the concept of “aeroelasticity” to estimate, model, emulate, simulate, and / or otherwise determine the dynamic response of an elastic structure or its components to a fluid. As illustrated by Collar's triangle, the concept of “aeroelasticity” may reflect the coupling effects among inertial forces, elastic forces, and aerodynamic forces when an elastic body (e.g., one or more rotor blades 112, rotor 108, and / or tower 102) is exposed to a fluid flow (e.g., a wind resource).

[0038] In one embodiment, the aeroelastic model 402 of the estimator utilizing aeroelastic concepts can include aerodynamic algorithms configured to determine aerodynamic loads developed in response to the flow pattern of the wind against the elastic body. The aerodynamic loads can be modeled, for example, via various computational fluid dynamics approaches. The aeroelastic model 402 can also include structural dynamics algorithms, such as finite element models, configured to model the dynamic response of the elastic body based on the structural mechanics of the elastic body. Thus, the aerodynamic and structural dynamics algorithms of the aeroelastic model 402 can be used together to model the fluid flow parameters (e.g., outputs) necessary to produce the observed or projected behavior (e.g., inputs) of the elastic body. In additional embodiments, the aerodynamic and structural mechanics algorithms of the aeroelastic model 402 may be utilized in tandem to model the expected behavior (e.g., output) of an elastic body in response to observed or predicted fluid flow (e.g., input). The aeroelastic model 402 may include at least one algorithm directed to, for example, a boundary element method, a Glauert correction factor, an induced velocities with dynamic inflow model, a blade deformation, quasi-static blade twist, and / or other suitable algorithms.

[0039] As particularly depicted in FIGS. 7 and 8 , in one embodiment, each aeroelastic estimator 322 of the plurality of aeroelastic estimators 320 may be tuned to a different presumptive wind condition (e.g., a nominal wind condition (WC), WC-1, WC-2, WC-3, ... WC-n). The different presumptive wind conditions may include, for example, a nominal wind condition, a gusty condition (e.g., WC-1), a shear condition (e.g., WC-2), a bare condition (e.g., WC-3), a turbulent condition (e.g., WC-4), and / or other wind conditions representing deviations from the nominal wind condition that may affect the operation of the wind turbine 100. In other words, each aeroelastic estimator 322 may determine parameters of the fluid flow (e.g., wind) that would result in the observed or predicted operation of the wind turbine 100 if the fluid flow were experiencing the presumptive wind condition to which each aeroelastic estimator 322 is tuned.

[0040] As a non-limiting example, one aeroelastic estimator 322 of the plurality of aeroelastic estimators 320 may estimate that wind impinging on wind turbine 100 has shear conditions and model the inertial, elastic, and aerodynamic forces of wind turbine interactions under the estimated shear conditions. It should be understood that the same calculations may be accomplished in parallel (e.g., simultaneously) for each additional estimated wind condition by each additional aeroelastic estimator 322 of the plurality of aeroelastic estimators 320 (e.g., via a multi-core architecture). Furthermore, it should be understood that for a given operation of the wind turbine as reflected by current operating data 304, different estimated wind conditions for each aeroelastic estimator 322 of the plurality of aeroelastic estimators 320 may result in different output of fluid flow parameters from each aeroelastic estimator 322.

[0041] In one embodiment, the wind classification module 214 may generate a plurality of wind descriptive parameters 404 from each aeroelastic estimator 322 based at least in part on the current operational dataset 304. Thus, each of the plurality of wind descriptive parameters 404 may describe characteristics of the wind impacting the wind turbine 100 that would result in the operational dataset 304 if the wind demonstrated the estimated wind conditions for which the corresponding aeroelastic estimator 322 is tuned. Each of the plurality of wind descriptive parameters 404 may include, for example, parameters corresponding to wind speed, vertical shear, horizontal shear, vertical misalignment, horizontal misalignment, turbulence, and / or any other parameters describing the wind impacting the wind turbine 100.

[0042] In one embodiment, the wind classification module 214 may determine the current aerodynamic conditions 302 from the plurality of wind descriptive parameters 404 generated by the first plurality of aeroelastic estimators 320. It should be appreciated that the wind classification module 214 may determine the current aerodynamic conditions 302 based on a designated portion 406 of the plurality of wind descriptive parameters 404.

[0043] For example, in one embodiment, each aeroelastic estimator 322 of the first plurality of aeroelastic estimators 320 may include at least one aeroelastic model 402. In an additional embodiment, each aeroelastic estimator 322 of the first plurality of aeroelastic estimators 320 may include at least one filtering algorithm 408. The one or more aeroelastic models 402 of each aeroelastic estimator 322 may be configured to model the behavior of the wind turbine 100 as a multibody system of flexible structures. As such, the aeroelastic models 402 may be utilized to derive a resultant aerodynamic state 410 of the wind resource deploying the current operating dataset 304 in the presence of estimated wind conditions corresponding to the tuning of the aeroelastic models 402. Thus, the wind classification module 214 may determine 411 a plurality of wind description parameters 404 corresponding to the aerodynamic states 410 derived by one or more aeroelastic models 402 of each aeroelastic estimator 322 of the first plurality of aeroelastic estimators 320.

[0044] 6, to determine the specified portion 406 of the plurality of wind-describing parameters 404, the controller 200 may model a predicted operational response of the wind turbine 100 to one potential aerodynamic condition 412 of a plurality of potential aerodynamic conditions 414 of the wind resource. Based on the predicted operational response, the controller 200 may generate a predicted operational dataset 416. The predicted operational dataset 416 describes the predicted operation of the wind turbine 100 if the wind turbine 100 were to encounter the one potential aerodynamic condition 412.

[0045] It should be appreciated that the plurality of potential aerodynamic conditions 414 may correspond to the environmental operating envelope of the wind turbine 100. Thus, a potential aerodynamic condition 412 may be a wind condition that falls within the environmental operating envelope of the wind turbine, and therefore may be a wind condition that the wind turbine 100 may encounter. A plurality of potential wind parameters may be utilized to describe a potential aerodynamic condition 412 of the plurality of potential aerodynamic conditions 414.

[0046] As illustrated in step (b), in one embodiment, each aeroelastic estimator 322 of the first plurality of aeroelastic estimators 320 may be utilized by controller 200 to generate a plurality of predicted wind-descriptive parameters 418 based on the predicted operating data set 416. It should be appreciated that for a given predicted operating data set 416, different estimated wind conditions for each aeroelastic estimator 322 may result in different output of predicted wind-descriptive parameters 418.

[0047] As illustrated in step (c), in an embodiment, a portion 420 of the plurality of predicted wind-descriptive parameters 418 may be determined that replicates a potential aerodynamic state of the plurality of potential aerodynamic states 414. For example, the various individual predicted wind-descriptive parameters generated by the first plurality of aeroelastic estimators 320 based on the predictive operating dataset 416 may be compared to the multiple potential wind parameters that describe the potential aerodynamic state 412 modeled by the controller in step (a). This comparison may reveal which of the various individual predicted wind-descriptive parameters of the plurality of predicted wind-descriptive parameters 418 describe the potential aerodynamic state.

[0048] As indicated at 422, in an embodiment, step (d) may include associating a portion of the plurality of predicted wind-descriptive parameters 418 corresponding to one potential aerodynamic state 412 of the plurality of potential aerodynamic states 414 with the predicted operating dataset 416. In other words, a relationship may be established between a particular portion of the plurality of predicted wind-descriptive parameters 418 that describes one potential aerodynamic state 412 and the predicted operating dataset 416 from which the plurality of predicted wind-descriptive parameters 418 were generated.

[0049] As illustrated in step (e), in an embodiment, steps (a) through (d) may be repeated for the remainder 424 of the plurality of potential aerodynamic states 414. It should be appreciated that repeating steps (a) through (d) may facilitate determining correlations between portions of the plurality of predicted wind descriptive parameters 418 and the predicted operational datasets 416 corresponding to the various remaining 424 potential aerodynamic states 414. Thus, correlations may be determined across the environmental operating envelope of the wind turbine 100.

[0050] As shown in step (f), in one embodiment, controller 200 may generate an operational-response signature data set 426. The operational-response signature data set 426 may include the predicted operational data set 416 and the corresponding portion 420 of the plurality of predicted wind-descriptive parameters 418 for each of the plurality of potential aerodynamic states 414. In other words, a combination of a particular predicted operational data set 416 and a particular corresponding portion 420 of the plurality of predicted wind-descriptive parameters 418 may indicate a particular aerodynamic state. Furthermore, operational-response signature data set 426 may facilitate identification of a wind state corresponding to one of the estimated wind states (e.g., nominal WC, WC-1, WC-2, WC-3, ... WC-n) based on the particular corresponding portion 420 of the plurality of predicted wind-descriptive parameters 418 generated by first plurality of aeroelastic estimators 320 based on the particular predicted operational data set 416. The motion response signature data set 426 may be assembled by the controller 200 into a database, a look-up table, and / or a graphical representation.

[0051] It should be appreciated that the operational response signature dataset 426 may facilitate determining the current aerodynamic state 302 of the wind resource (including current wind conditions such as nominal, gust, shear, veer, turbulence, etc.) based on the plurality of wind-descriptive parameters 404 generated by the first plurality of aeroelastic estimators 320 based on the current operational dataset 304. It should further be appreciated that steps (a)-(f) may, in one embodiment, be accomplished offline via the controller 200.

[0052] As depicted at 428, during operation, in an embodiment, the wind classification module 214 may determine a designated portion 406 of the plurality of wind descriptive parameters 404 based on the motion response signature dataset 426, the plurality of wind descriptive parameters 404, and the current motion dataset 304. In other words, the portion of the plurality of wind descriptive parameters 404 indicative of the current aerodynamic state 302 may be identified / designated based on the relationship between the plurality of wind descriptive parameters 404 indicated by the motion response signature dataset 426 and the current motion dataset 304.

[0053] In one embodiment, the wind classification module 214 may include at least one machine learning algorithm 430. Thus, to determine the specified portion 406 of the plurality of wind descriptive parameters 404, the controller 200 may implement the one or more machine learning algorithms 430 in the wind classification module 214 to generate the operational response signature dataset 426. The machine learning algorithm 430 may utilize known machine learning models, such as Gaussian Process Models, Random Forest Models, Neural Networks, and / or Support Vector Machines. It should be understood that the one or more machine learning algorithms 430 may be trained offline / online via steps (a)-(f).

[0054] As depicted at 432, one or more machine learning algorithms 430 may, in embodiments, be implemented to determine the designated portion 406 of the plurality of wind descriptive parameters 404 corresponding to the current motion dataset 304. For example, in embodiments, the one or more machine learning algorithms 430 may generate and / or utilize the motion response signature dataset 426 to determine the designated portion 406.

[0055] As depicted at 434, in an embodiment, one or more machine learning algorithms 430 may be implemented by controller 200 to determine current aerodynamic conditions 302 based on designated portion 406 of plurality of wind-descriptive parameters 404. For example, in an embodiment, one or more machine learning algorithms 430 may generate and / or utilize motion response signature dataset 426 to determine designated portion 406, and then determine current aerodynamic conditions 302 based on designated portion 406.

[0056] 5, 7, and 8, in one embodiment, the configuration intelligence module 216 may be utilized by the controller 200 to generate an estimator configuration 306 and a predictive control configuration 308. The estimator configuration may include, for example, at least a process-noise gain (Q) and a sensor-noise gain (R). The predictive control configuration may include, for example, at least gains corresponding to tracking weights 334, slack weights 336, and constraint limits 338.

[0057] Thus, in embodiments, configuration intelligence module 216 may select a gain tuning 324 of multiple gain tunings 326 based on current aerodynamic conditions 302 to generate estimator configuration 306 and predictive control configuration 308. In such embodiments, each gain tuning of multiple gain tunings 326 may be configured to modify estimator configuration 306 and / or predictive control configuration 308.

[0058] 7 and 8, each gain tuning of the plurality of gain tunings 326 may correspond to a desired performance / operational response of the wind turbine 100 when encountering a particular wind condition. For example, the plurality of gain tunings 326 may include a gain tuning corresponding to each of the estimated wind conditions (e.g., nominal WC, WC-1, WC-2, WC-3, ... WC-n) to which the plurality of aeroelastic estimators 320 may be adjusted.

[0059] 7 illustrates an embodiment in which the designated portion 406 of the plurality of wind description parameters 404 indicates a current aerodynamic condition 302 that matches a nominal wind condition for the wind resource affecting the wind turbine 100. As such, the configuration intelligence module 214 may select a gain tuning 324 that is configured as a nominal-WC gain tuning. In other words, in an embodiment in which the wind classification module 214 determines that the wind resource affecting the wind turbine 100 has a nominal condition, the configuration intelligence module 216 may select a gain tuning 324 that optimizes the performance of the wind turbine 100 for the nominal wind conditions.

[0060] 8 depicts an embodiment in which the designated portion 406 of the plurality of wind description parameters 404 indicates a current aerodynamic condition 302 that is consistent with a first wind condition (WC-1) (e.g., any of gusty, shear, bear, turbulent, etc.). As such, the configuration intelligence module 214 may select the gain tuning 324 configured as the WC-1 gain tuning. In other words, in an embodiment in which the wind classification module 214 determines that the wind resource impacting the wind turbine 100 has a first wind condition, the configuration intelligence module 216 may select the gain tuning 324 that optimizes the performance of the wind turbine 100 for the determined current aerodynamic condition 302.

[0061] It should be appreciated that configuration intelligence module 216 may switch between gain tunings of plurality of gain tunings 326 in response to changes in determination of current aerodynamic state 302 by wind classification module 214. For example, in an embodiment, configuration intelligence module 216 may detect a difference 328 between current aerodynamic state 302 and immediate-past aerodynamic state 330. Difference 328 may require a gain transition between various gain tunings of plurality of gain tunings 326.

[0062] In embodiments where a gain transition may be required, the configuration intelligence module 216 may implement a transition algorithm 332. The transition algorithm 332 may correspond to the current aerodynamic conditions 302. In other words, the transition algorithm 332 may be tailored to the current aerodynamic conditions 302. For example, the transition algorithm 332 may implement a fast blend of gain transitions for sudden events / wind conditions, such as gusts of wind. However, the transition algorithm 332 may perform slower blending for events / wind conditions, such as shear or bear. In this manner, the transition algorithm 332, in embodiments, may facilitate a smooth transition between the previous gain tuning and the selected gain tuning 324 corresponding to the current aerodynamic conditions 302.

[0063] 4, 5, and 9, to generate the estimator configuration 306 and the predictive control configuration 308, the controller 200 may, in embodiments, also utilize an operating state module 222. The operating state module 222 may determine a current operating state 340 of the wind turbine 100. The current operating state 340 may, for example, indicate that the current operation 305 of the wind turbine 100 is consistent with the nominal operation of the wind turbine 100. However, in additional embodiments, the current operating state 340 may indicate that the current operation 305 has degraded relative to the nominal operation of the wind turbine 100. In further embodiments, the current operating state 340 may indicate the presence of a fault condition affecting a component and / or sensor signal of the wind turbine 100.

[0064] Determining the current operating state 340 may be accomplished by performance of the process 500 by the operating state module 222. In an embodiment, the current operating state 340 may be determined, at least in part, based on the current operating dataset 304. In such an embodiment, the current operating dataset 304 may include a plurality of output signals 502 from one or more sensor systems 154 of the wind turbine 100. Further, the current operating dataset 304, in an embodiment, may include at least one control signal 504 from the controller 200.

[0065] In embodiments in which system 300 employs operating conditions module 222, configuration intelligence module 216 may generate estimator configuration 306 for turbine estimator module 218 based at least in part on current aerodynamic conditions 302 and current operating conditions 340. In additional embodiments, configuration intelligence module 216 may generate predictive control configuration 308 for predictive control module 220 based at least in part on current aerodynamic conditions 302 and current operating conditions 340.

[0066] In an embodiment, operating conditions module 222 may include a second plurality of aeroelastic estimators 342. Each aeroelastic estimator 344 of the second plurality of aeroelastic estimators 342 may be tuned to a different estimated fault / abnormal condition of wind turbine 100. In other words, each aeroelastic estimator 344 of the second plurality of aeroelastic estimators 342 may estimate that current operating data 304 includes output signals 502 and / or control signals 504 that are a result of the estimated fault condition to which the particular aeroelastic estimator 344 is tuned.

[0067] For example, in one embodiment, at least one aeroelastic estimator 344 may be tuned to describe the operation of wind turbine 100 based on an estimation that wind turbine 100 is experiencing a blade pitch anomaly. The blade pitch anomaly may include a pitch system condition such as a frozen pitch angle, a sluggish actuator response, excessive bearing friction, and / or other similar conditions that may affect the ability of the pitch system to orient rotor blades 112 at a desired pitch angle.

[0068] In additional embodiments, the at least one aeroelastic estimator 344 can be adjusted to describe the operation of the wind turbine 100 based on an estimation that the wind turbine 100 is experiencing a sensor anomaly. The sensor anomaly can include biases, excessive noise, and / or failures in the sensor system 154 and / or the operational sensors 158. For example, the sensor anomaly may be indicated by a lack of signal and / or poor signal quality / resolution from one or more sensors. In additional embodiments, a sensor failure may be indicated by a sensor output being out of range and / or a missing azimuth signal. Additional sensor faults include, but are not limited to, a zero-calibration-low fault, a strain sensor drift, a temperature sensor failure, and an additive fault, a multiplicative fault, an output stuck fault, and / or a slow drifting fault.

[0069] In a further embodiment, at least one aeroelastic estimator 344 can be tuned to describe the operation of wind turbine 100 based on an estimation that wind turbine 100 is experiencing degraded / suboptimal performance. Such degradation may be due to, for example, ice formation, pitch angle offsets, a blade imbalance, a degraded airfoil, a wake effect, a wind shadow, and / or other conditions that may cause the operation of wind turbine 100 to deviate from nominal operation. It should be appreciated that individual aeroelastic estimators of second plurality of aeroelastic estimators 342 can be tuned to estimate the presence of different sources of degradation.

[0070] In one embodiment, the operating conditions module 222 may generate the plurality of descriptive operating parameters 506 from the second plurality of aeroelastic estimators 342. The generation of the plurality of descriptive operating parameters 506 may be based, at least in part, on the current operating dataset 304. In an embodiment, as depicted at 507, the operating state module 222 may determine the current operating state 340 of the wind turbine 100 based on a designated portion 508 of the plurality of descriptive operating parameters 506.

[0071] As particularly depicted in FIG. 9 , to determine the specified portion 508 of the plurality of descriptive operating parameters 506, the controller 200 may model predicted sensor outputs 510 from the sensor system 154. The predicted sensor outputs 510 may be modeled for predicted operating conditions 514 under one potential fault condition of the plurality of potential fault conditions 512. The second plurality of aeroelastic estimators 342 may then be utilized to generate the plurality of predicted descriptive operating parameters 516 based on a predicted operational dataset 518 corresponding to the predicted operating conditions 514. As depicted at 520, a portion of the plurality of predicted descriptive operating parameters 516 may be determined that replicates the projected plurality of sensor outputs 510. As depicted at 522, the portion of the plurality of predicted descriptive operating parameters 516 may be correlated to the predicted sensor outputs corresponding to the one potential fault condition of the plurality of potential fault conditions 512. As depicted at 524 , the steps may be repeated for the remaining portion of the plurality of potential fault conditions 512 .

[0072] In an embodiment, the operating condition module 222 may implement at least one machine learning algorithm to generate a fault signature dataset 526. The fault signature dataset may include a plurality of predicted sensor outputs 510 at each potential fault state of a plurality of potential fault states 512 and a corresponding portion of a plurality of predicted descriptive operating parameters 516. In an embodiment, as depicted at 528, one or more machine learning algorithms may then be implemented to determine a designated portion 508 of the plurality of descriptive operating parameters 506 based on the fault signature dataset 526 and the current operating dataset 304. It should be understood that one or more machine learning algorithms may also be implemented to determine a current operating state 340 based on the designated portion 508 of the plurality of descriptive operating parameters 506, as depicted at 530.

[0073] In embodiments in which the system 300 employs an operating condition module 222, the configuration intelligence module 216 may select the gain tunings 324 based on the current aerodynamic conditions 302 and the current operating conditions 340. In such embodiments, each gain tuning 324 may be configured to modify at least one of the estimator configuration 306 and the predictive control configuration 308 based on the current aerodynamic conditions 302 of the wind resource and the current operating conditions 340 of the wind turbine 100.

[0074] To select a gain tuning 324 of the plurality of gain tunings 326, in one embodiment, a prioritization schedule 346 may be determined for the plurality of gain tunings 326. Thus, as depicted at 348, configuration intelligence module 216 may select gain tunings 324 according to prioritization schedule 346 based on current aerodynamic conditions 302 and current operating conditions 340. For example, in one embodiment, prioritization schedule 346 may prioritize gain tunings 326 associated with throttling actions in response to a particular fault / anomaly over gain tunings 326 configured to optimize wind turbine performance when particular wind conditions are encountered.

[0075] 3 and 4 , the system 300, in one embodiment, may include a feedback evaluation module 224. The feedback evaluation module 224 may receive at least one feedback signal 350 from the turbine estimator module 218. The one or more feedback signals 350 may indicate an emulated operation of the wind turbine 100 in response to the aerodynamic conditions 302 and / or the current operating conditions 340. In other words, the one or more feedback signals 350 may correspond to a digital representation of the physical wind turbine 100 when encountering the current aerodynamic conditions 302 and / or the current operating conditions 340. Then, as depicted at 352, the controller 200, in an embodiment, may modify at least one gain tuning 324 of the plurality of gain tunings 326 of the configuration intelligence module 216 based on the one or more feedback signals 350.

[0076] 4 , in additional embodiments, wind classification module 214 and / or operating condition module 222 may receive one or more feedback signals 350 from turbine estimator module 218. In response to receiving feedback signal 350, wind classification module 214 may modify plurality of wind descriptive parameters 404 and / or designated portion 406 of plurality of wind descriptive parameters 404. Similarly, in embodiments, operating condition module 222 may modify plurality of descriptive operating parameters 506 and / or designated portion 508 of plurality of descriptive operating parameters 506 based on feedback signal 350.

[0077] In an embodiment, controller 200 may be configured to perform multiple parallel calculations to determine or generate estimator configuration 306, predictive control configuration 308, control initial state 310, predicted performance 312 of wind turbine 100, and setpoint 316. For example, controller 200 may be configured to leverage the computational benefits of advanced vector extension instructions. Accordingly, processor 206 of controller 200 may have a multi-core architecture. Utilizing a multi-core architecture may facilitate parallel execution of multiple advanced vector extension instructions.

[0078] In an embodiment, turbine estimator module 218 may calculate a continuous data stream indicative of the structural condition of wind turbine 100 at any given moment. For example, in an embodiment, turbine estimator module 218 may be configured according to estimator configuration 306 and receive current operating data 304. In response, turbine estimator module 218 may emulate the structural condition of wind turbine 100. It should be appreciated that the structural condition of wind turbine 100 may correspond to the current state of wind turbine 100 and thus establish control initial state 310, from which predictive control module 220 may predict the performance of the wind turbine over prediction horizon 314. In an embodiment, the prediction interval may be less than 10 seconds (e.g., 8 seconds).

[0079] In an embodiment, the system 300 may utilize the predictive control module 220 to determine predicted performance 312 of the wind turbine 100 over a prediction horizon 314 and generate setpoints 316 for the actuators 160 based on the predicted performance 312. In this manner, the predictive control module may include a mathematical representation (e.g., a model) of the system dynamics of the wind turbine 100 in a linearized form. As such, the mathematical representation may be derived from first principles or by other means (e.g., data driven models). In this manner, the predictive control module 220 may be configured to determine the results of various high fidelity simulation scenarios describing the operation of the wind turbine 100.

[0080] Starting from the control initial state 310, the predictive control module 220, in one embodiment, may propagate the mathematical representation of the wind turbine 100 for future times (e.g., predictive interval 314). For example, the predictive control module 220 may simulate / emulate the predicted response of the wind turbine 100 to the current aerodynamic state 302 and / or current operating conditions 340. In other words, the predictive control module 220 may utilize model-based techniques to continuously emulate the operation of the wind turbine in response to actual conditions affecting the wind turbine 100 as reflected by the current operating data 304. It should be understood that the predictive control module 220 may be configured to digitally replicate every aspect of the wind turbine 100.

[0081] It should also be appreciated that the predictive control module 220 may utilize the predictive capabilities of the mathematical representation / model to optimize control actions (e.g., pitch and / or torque) over the prediction horizon 314. The optimization may specify constraints to be respected during the prediction horizon 314. The constraints may include, for example, blade loads, tower loads, pitch amplitude and rate constraints, torque amplitude and rate constraints, pitch travel, and / or other similar constraints. Furthermore, the optimization may have a cost function configured to minimize penalties for deviations from performance targets and penalties for control actions. The solution of the optimization may provide an optimal value for the setpoint 316 over the prediction horizon 314. The controller 200 may update the control initial conditions 310 and solve a new optimization for the current time. In this manner, the optimization process is continuously repeated. In other words, the optimization problem is initialized and solved at the current time, and a new optimization process is started with a receiving horizon approach.

[0082] Moreover, those skilled in the art will recognize the interchangeability of various features from different embodiments. Similarly, the various method steps and features described, and other known equivalents to each such method and feature, can be mixed and matched by those skilled in the art to construct additional systems and techniques in accordance with the principles of the present disclosure. Of course, it should be understood that not necessarily all such objects or advantages may be achieved in accordance with any particular embodiment. Thus, for example, those skilled in the art will recognize that the systems and techniques described herein can be embodied or implemented in a way that achieves or optimizes one advantage or advantages as taught herein, without necessarily achieving other objects or advantages that may be taught or suggested herein.

[0083] This specification uses examples to disclose the invention, including the best mode, and also to enable any person skilled in the art to practice the invention, including making and using any devices or systems, and performing any methods incorporated therein. The patentable scope of the invention is defined by the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they include structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements that do not differ substantially from the literal language of the claims.

[0084] Further aspects of the invention are provided by the subject matter of the following clauses. [Embodiment 1] 1. A method of controlling a wind turbine, the wind turbine including a controller, the method comprising: determining, via a wind classification module of the controller, a current aerodynamic state of the wind resource based at least in part on a wind turbine current operation dataset indicative of a current operation of the wind turbine; generating, via a configuration intelligence module of the controller, an estimator configuration for a turbine estimator module based at least in part on the current aerodynamic conditions; emulating operation of a wind turbine via the turbine estimator module of the controller to generate an initial control state for a predictive control module, the initial control state including modeled current operating states of a plurality of components of the wind turbine; generating, via the configuration intelligence module, a predictive control configuration for the predictive control module based at least in part on the current aerodynamic conditions; determining, via the predictive control module of the controller, a predicted performance of the wind turbine over a prediction interval based on the control initial state and the predictive control configuration; generating, via the predictive control module, a setpoint for at least one actuator of the wind turbine based on the predicted performance; and influencing an operating state of the wind turbine via the at least one actuator according to the setpoint. [Embodiment 2] 2. The method of embodiment 1, wherein the wind classification module comprises a first plurality of aeroelastic estimators, each aeroelastic estimator being tuned to a different estimated wind condition, and wherein determining the current aerodynamic condition further comprises generating, via the wind classification module, a plurality of wind descriptive parameters from each aeroelastic estimator based at least in part on the current operating dataset, and determining, via the wind classification module, the current aerodynamic condition based on a designated portion of the plurality of wind descriptive parameters. [Embodiment 3] The step of determining the designated portion of the plurality of wind descriptive parameters comprises: a) modeling, via the controller, a predicted operational response of a wind turbine to one potential aerodynamic state of a plurality of potential aerodynamic states of a wind resource to generate a predicted operational data set; b) generating, via each aeroelastic estimator of the first plurality of aeroelastic estimators, a plurality of predicted wind descriptive parameters based on the predicted operating data set; c) determining a subset of a plurality of predicted wind-descriptive parameters that represent one potential aerodynamic state; d) correlating a portion of said plurality of predicted wind-descriptive parameters with a predicted operational data set corresponding to potential aerodynamic conditions; e) repeating steps a) to d) for the remainder of the plurality of potential aerodynamic states, the plurality of potential aerodynamic states corresponding to the environmental operating envelope of the wind turbine; and f) generating, via said controller, a motion response signature data set comprising a predicted motion data set for each potential aerodynamic state of said plurality of potential aerodynamic states and a corresponding portion of a plurality of predicted wind-descriptive parameters; A method as described in embodiment 1 or 2, comprising determining, via the wind classification module, the designated portion of the plurality of wind descriptive parameters based on the operational response signature dataset, the plurality of wind descriptive parameters, and the current operational dataset. [Embodiment 4] The step of determining the designated portion of the plurality of wind description parameters further comprises: Implementing at least one machine learning algorithm within the wind classification module to generate a motion response signature dataset; Implementing at least one machine learning algorithm to determine a designated portion of a plurality of wind descriptive parameters corresponding to the current operating data set; implementing the at least one machine learning algorithm to determine the current aerodynamic state based on the specified portion of the plurality of wind-descriptive parameters; 10. The method of any preceding embodiment, comprising: [Embodiment 5] wherein each aeroelastic estimator of the first plurality of aeroelastic estimators comprises at least one aeroelastic model and at least one filtering algorithm, and wherein the at least one aeroelastic model of each aeroelastic estimator is configured to model the behavior of the wind turbine as a flexible structure multibody system, and to generate the plurality of wind descriptive parameters; deriving, via the at least one aeroelastic model, a resultant aerodynamic state of a wind resource developing a current operating dataset in the presence of estimated wind conditions corresponding to tuning of the at least one aeroelastic model; determining a plurality of wind-descriptive parameters corresponding to the resultant aerodynamic state derived by at least one aeroelastic model of each aeroelastic estimator of the first plurality of aeroelastic estimators. [Embodiment 6] generating the estimator configuration and the predictive control configuration, 10. The method of any preceding embodiment, comprising selecting, via a configuration intelligence module, a gain tuning from a plurality of gain tunings based on current aerodynamic conditions, each gain tuning configured to modify at least one of the estimator configuration and the predictive control configuration. [Embodiment 7] The step of generating the estimator configuration and the predictive control configuration further comprises: determining, via an operational state module of the controller, a current operational state of the wind turbine based at least in part on the current operational dataset, the operational dataset further comprising a plurality of output signals from a sensor system of the wind turbine and at least one control signal from the controller; generating, via a configuration intelligence module of the controller, an estimator configuration of a turbine estimation module based at least in part on the current aerodynamic state and the current operating state; and generating, via the configuration intelligence module, a predictive control configuration of a predictive control module based at least in part on the current aerodynamic state and the current operating state. [Embodiment 8] the operating state module includes a second plurality of aeroelastic estimators, each aeroelastic estimator tuned to a different estimated fault condition of the wind turbine, and determining the current operating state further comprises: generating, via the operating conditions module, a plurality of descriptive operating parameters from the second plurality of aeroelastic estimators based at least in part on the current operating data set; 10. The method of any preceding embodiment, further comprising: determining, via the operating condition module, the current operating conditions of the wind turbine based on a designated portion of the plurality of descriptive operating parameters. [Embodiment 9] The step of determining the designated portion of the plurality of descriptive operating parameters comprises: a) modeling, via the controller, predicted sensor outputs from a sensor system under one potential fault condition of a plurality of potential fault conditions for a predicted operating state; b) generating, via the second plurality of aeroelastic estimators, a plurality of predicted descriptive operational parameters based on a predicted operational dataset corresponding to the predicted operational state; c) determining a portion of said plurality of predicted descriptive operating parameters that replicate said predicted plurality of sensor outputs; e) repeating steps a) through d) for the remainder of the plurality of potential fault conditions; and d) correlating a portion of the plurality of predicted descriptive operating parameters and predicted plurality of sensor outputs with a corresponding potential fault condition; Implementing at least one machine learning algorithm within the operational state module to generate a fault signature data set consisting of predicted sensor outputs and a corresponding portion of a plurality of predicted descriptive operational parameters under each potential fault condition; Implementing at least one machine learning algorithm to determine a designated portion of the plurality of descriptive operational parameters based on the fault signature dataset and the current operational dataset; Implementing at least one machine learning algorithm to determine the current operating state based on the designated portion of the plurality of descriptive operating parameters. [Embodiment 10] 10. The method of any preceding embodiment, wherein each aeroelastic estimator of the second plurality of aeroelastic estimators includes at least one aeroelastic model and at least one filtering algorithm, and wherein the at least one aeroelastic model of each aeroelastic estimator is configured to model the behavior of the wind turbine as a flexible multibody system. [Embodiment 11] selecting the gain tuning from the plurality of gain tunings, selecting, via the configuration intelligence module, a gain tuning based on current aerodynamic conditions and current operating conditions; 10. The method of any preceding embodiment, wherein each gain tuning is configured to modify at least one of the estimator configuration and the predictive control configuration based on a current aerodynamic state of the wind resource and a current operating state of the wind turbine. [Embodiment 12] selecting the gain tuning from the plurality of gain tunings, determining a prioritized schedule for the plurality of gain tunings; 10. The method of any preceding embodiment, further comprising: via the configuration intelligence module, selecting gain tuning according to the prioritization schedule based on the current aerodynamic state and the current operating state. [Embodiment 13] selecting the gain tuning from the plurality of gain tunings, receiving at least one feedback signal from the turbine estimator module via a feedback evaluation module of the controller, the at least one feedback signal indicative of emulated operation of the wind turbine in response to the aerodynamic conditions and the current operating conditions; 10. The method of any preceding embodiment, further comprising: modifying, via the controller, at least one gain tuning of a plurality of gain tunings of the configuration intelligence module based on the at least one feedback signal. [Embodiment 14] generating the estimator configuration and the predictive control configuration comprises: detecting, via the configuration intelligence module, a difference between a current aerodynamic state requiring a gain transition and a previous aerodynamic state; Implementing, via the configuration intelligence module, a transition algorithm corresponding to the current aerodynamic condition to smoothly transition between a previous gain tuning and a selected gain tuning corresponding to the current aerodynamic condition. [Embodiment 15] 10. The method of any preceding embodiment, wherein the estimator configuration includes at least a process noise gain and a sensor noise gain. [Embodiment 16] 10. The method of any preceding embodiment, wherein the predictive control configuration includes at least gains corresponding to tracking weights, slack weights, and constraint limits. [Embodiment 17] 10. The method of any preceding embodiment, further comprising performing, via the controller, a plurality of parallel computations to determine or generate the estimator configuration, the predictive control configuration, the control initial state, the predicted performance of the wind turbine, and the setpoint. [Embodiment 18] 10. The method of any preceding claim, wherein performing a plurality of parallel computations further comprises executing a plurality of advanced vector extension instructions in parallel. [Embodiment 19] receiving at least one feedback signal from the turbine estimator module via the wind classifier module and the operating conditions module, the at least one feedback signal indicative of emulated operation of the wind turbine in response to aerodynamic conditions and current operating conditions; and modifying, via the controller, at least one of the plurality of descriptive operating parameters, the designated portion of the plurality of descriptive operating parameters, the plurality of wind descriptive parameters, or the designated portion of the plurality of wind descriptive parameters based on the at least one feedback signal. [Embodiment 20] 10. The method of any preceding embodiment, wherein emulating operation of the wind turbine further comprises calculating, via the turbine estimation module, a continuous stream of data indicative of a structural condition of the wind turbine. [Explanation of symbols]

[0085] 100: Wind turbine 102: Tower 104: Support surface 106: Nacelle 108: Rotor 110: Hub 112: Rotor blades 116: Pitch axis 118: Generator 120: Pitch control mechanism 122: Rotor shaft 124: High speed shaft 126: Gearbox 128: Pitch drive motor 130: Pitch drive gearbox 132: Pitch drive pinion 134: Pitch bearing 136: Bed plate support frame 138: Yaw drive mechanism 140: Yaw bearing 144: Main bearing 154: Sensor system 158: Motion sensor 160: Actuator 200: Controller 202: Turbine controller 206: Processor 208: Memory device 210: Communication module 212: Sensor interface 214: Wind classification module 216: Configuration intelligence module 218: Turbine estimation module 220: Predictive Control Module 222: Operating State Module 224: Feedback Evaluation Module 300: System 302: Current Aerodynamic State 304: Current Operating Dataset 305: Current Operation 306: Estimator Configuration 308: Predictive Control Configuration 310: Control Initial State 312: Predicted Performance 314: Prediction Interval 316: Setpoint 318: Operating State 320, 322: Aeroelastic Estimator 324, 326: Gain Tuning 328: Difference 330: Previous / Past Aerodynamic State 332: Transition Algorithm 334: Tracking Weights 336: Slack Weights 338: Constraint Limits 340: Operating Conditions 342, 344: Aeroelastic Estimator 346: Prioritization Schedule 350: Feedback Signal 400: Process 402: Aeroelastic Model 404: Wind Description Parameters 406: Specified Part 408: Filtering algorithm 410: Aerodynamic states 412, 414: Potential aerodynamic states 416: Predicted operating data set 420: Part 424: Remaining 500 Process 502: Output signal 504: Control signal 506: Plurality of descriptive operating parameters 512: Plurality of potential fault states 514: Predicted operating conditions 518: Predicted operating data set

Claims

1. 1. A method of controlling a wind turbine, the wind turbine including a controller, the method comprising: determining, via a wind classification module of the controller, a current aerodynamic state of the wind resource based at least in part on a wind turbine current operation dataset indicative of a current operation of the wind turbine; generating, via a configuration intelligence module of the controller, an estimator configuration for a turbine estimator module based at least in part on the current aerodynamic conditions; emulating operation of the wind turbine via a turbine estimation module of the controller to generate initial control conditions for a predictive control module, the initial control conditions including modeled current operating conditions of a plurality of components of the wind turbine; generating, via the configuration intelligence module, a predictive control configuration for the predictive control module based at least in part on current aerodynamic conditions; determining, via the predictive control module of the controller, a predicted performance of the wind turbine over a prediction horizon based on the control initial state and the predictive control configuration; generating, via the predictive control module, a setpoint for at least one actuator of the wind turbine based on the predicted performance; and influencing an operating state of the wind turbine via the at least one actuator according to the setpoint.

2. The wind classification module comprises a first plurality of aeroelastic estimators, each aeroelastic estimator tuned to a different estimated wind condition, and the step of determining the current aerodynamic condition further comprises: generating, via the wind classification module, a plurality of wind-descriptive parameters from each aeroelastic estimator based at least in part on the current operational dataset; and determining, via the wind classification module, the current aerodynamic conditions based on a designated portion of the plurality of wind descriptive parameters.

3. determining the designated portion of the plurality of wind description parameters; a) modeling, via the controller, a predicted operational response of the wind turbine to one potential aerodynamic state of a plurality of potential aerodynamic states of the wind resource to generate a predicted operational data set; b) generating, via each aeroelastic estimator of the first plurality of aeroelastic estimators, a plurality of predicted wind-descriptive parameters based on the predicted operational data set; c) determining a portion of said plurality of predicted wind-descriptive parameters that reproduce said one potential aerodynamic state; d) associating a portion of the predicted plurality of wind-descriptive parameters with the predicted operational data set corresponding to the potential aerodynamic state; e) repeating steps a) through d) for the remainder of the plurality of potential aerodynamic states; and f) generating, via said controller, a motion response signature data set including a predicted motion data set in each potential aerodynamic state of said plurality of potential aerodynamic states and said corresponding portion of said plurality of predicted wind-descriptive parameters; and determining, via the wind classification module, the designated portion of the plurality of wind descriptive parameters based on the motion response signature dataset, the plurality of wind descriptive parameters, and the current motion dataset.

4. The step of determining the designated portion of the plurality of wind description parameters further comprises: implementing at least one machine learning algorithm within the wind classifier module to generate the operational response signature data set; implementing the at least one machine learning algorithm to determine a designated portion of a plurality of wind-descriptive parameters corresponding to the current operating dataset; and implementing the at least one machine learning algorithm to determine the current aerodynamic state based on the specified portion of the plurality of wind-descriptive parameters.

5. wherein each aeroelastic estimator of the first plurality of aeroelastic estimators includes at least one aeroelastic model and at least one filtering algorithm, the at least one aeroelastic model of each aeroelastic estimator configured to model the behavior of the wind turbine as a flexible multibody system, and generating the plurality of wind descriptive parameters further comprises: deriving, via the at least one aeroelastic model, a resultant aerodynamic state of a wind resource that evolves the current operational dataset in the presence of estimated wind conditions corresponding to tuning of the at least one aeroelastic model; and determining a plurality of wind-descriptive parameters corresponding to the resulting aerodynamic state derived by at least one aeroelastic model of each aeroelastic estimator of the first plurality of aeroelastic estimators.

6. generating the estimator configuration and the predictive control configuration further comprises:

6. The method of claim 2, further comprising selecting, via the configuration intelligence module, a gain tuning from a plurality of gain tunings based on the current aerodynamic state, each gain tuning configured to change at least one of an estimator configuration and a predictive control configuration.

7. generating the estimator configuration and the predictive control configuration further comprises: determining, via the operational state module of the controller, a current operational state of the wind turbine based at least in part on the current operational dataset, the operational dataset further comprising a plurality of output signals from a sensor system of the wind turbine and at least one control signal from the controller; generating, via the configuration intelligence module of the controller, an estimator configuration for a turbine estimator module based at least in part on the current aerodynamic state and the current operating state; and generating, via the configuration intelligence module, a predictive control configuration for a predictive control module based at least in part on the current aerodynamic conditions and the current operating conditions.

8. the operational state module includes a second plurality of aeroelastic estimators, each aeroelastic estimator tuned to a different estimated fault condition of the wind turbine, and determining the current operational state further comprises: generating, via the operating conditions module, a plurality of descriptive operating parameters from a second plurality of aeroelastic estimators based at least in part on the current operating data set; and determining, via the operational state module, a current operational state of the wind turbine based on a designated portion of the plurality of descriptive operational parameters.

9. determining the designated portion of the plurality of descriptive operating parameters; a) modeling, via the controller, predicted values ​​of a plurality of sensor outputs from the sensor system under one potential fault condition of a plurality of potential fault conditions for a predicted operating state; b) generating, via the second plurality of aeroelastic estimators, a plurality of predicted descriptive operational parameters based on predicted operational data sets corresponding to the predicted operational states; c) determining a portion of the predicted plurality of descriptive operating parameters that replicate the predicted plurality of sensor outputs; d) associating the predicted sensor outputs and a portion of the predicted descriptive operating parameters with a corresponding potential fault condition; e) repeating steps a) through d) for the remainder of the plurality of potential fault conditions; Implementing at least one machine learning algorithm within the operational state module to generate a fault signature data set comprising predicted values ​​of a plurality of sensor outputs and a corresponding portion of a plurality of predicted descriptive operational parameters under each potential fault condition; implementing the at least one machine learning algorithm to determine a designated portion of a plurality of descriptive operational parameters based on the fault signature dataset and the current operational dataset; and implementing the at least one machine learning algorithm to determine the current operating state based on the specified portion of the plurality of descriptive operating parameters.

10. 9. The method of claim 8, wherein each aeroelastic estimator of the second plurality of aeroelastic estimators includes at least one aeroelastic model and at least one filtering algorithm, and wherein the at least one aeroelastic model of each aeroelastic estimator is configured to model the behavior of the wind turbine as a flexible multibody system.

11. The step of selecting the gain tuning from the plurality of gain tunings further comprises:

10. The method of claim 8 or 9, comprising selecting, via the configuration intelligence module, the gain tunings based on the current aerodynamic conditions and the current operating conditions, each gain tuning configured to change at least one of the estimator configuration and the predictive control configuration based on the current aerodynamic conditions of the wind resource and the current operating conditions of the wind turbine.

12. The step of selecting the gain tuning from the plurality of gain tunings further comprises: determining a prioritized schedule for the plurality of gain tunings; and selecting, via the configuration intelligence module, the gain tuning according to a prioritization schedule based on the current aerodynamic state and the current operating state.

13. The step of selecting the gain tuning from the plurality of gain tunings further comprises: receiving at least one feedback signal from the turbine estimator module via a feedback evaluation module of the controller, the at least one feedback signal indicative of emulated operation of the wind turbine in response to the aerodynamic conditions and the current operating conditions; 13. The method of claim 12, further comprising: altering, via the controller, at least one gain tuning of the plurality of gain tunings of the configuration intelligence module based on the at least one feedback signal.

14. The step of generating the estimator configuration and the predictive control configuration further comprises: detecting, via the configuration intelligence module, a difference between the current aerodynamic state and a previous aerodynamic state requiring a gain transition; and implementing, via the configuration intelligence module, a transition algorithm corresponding to the current aerodynamic state to smoothly transition between a previous gain tuning and a selected gain tuning corresponding to the current aerodynamic state.

15. 15. The method according to any one of claims 6 to 14, wherein the estimator configuration includes at least a process noise gain and a sensor noise gain.

Images