Method and system for predictive control of wind turbine

Abstract

To provide a method for controlling a wind turbine having a plurality of actuators.SOLUTION: A method comprises receiving operating data (366) of a wind turbine and determining an operating state of the wind turbine. Further, the method includes an intensity calculation module (372) for calculating secondary load parameters (374) from loads (375). The method comprises optimizing a cost function over a finite optimization time period, subject to constraints, to determine an optimal trajectory (391) including commands for actuators. Finally, the method includes controlling the actuators using the first commands of the determined optimal trajectory. The present disclosure relates to a wind turbine controller (360) configured to implement such a method.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present disclosure relates to wind turbines, and more particularly, to methods and systems for controlling a wind turbine using a prediction controller.

Background Art

[0002] Modern wind turbines are commonly used to supply power to an electrical grid. This type of wind turbine generally includes a tower and a rotor disposed on the tower. The rotor typically includes a hub and a plurality of blades, and rotates under the influence of the wind hitting the blades. This rotation generates a moment, which is usually transmitted directly to the generator through the rotor shaft ( "directly driven" or "gearless") or transmitted using a gearbox. Thus, the generator can generate electricity and supply the electricity to the power grid.

[0003] Variable speed wind turbines are generally controlled by changing the torque of the generator and the pitch angle of the blades. Generally, one or more control systems are arranged to determine the actuator signals for changing the torque of the generator and the pitch angle. Generally, the wind turbine is controlled to capture maximum power from the wind, and maximum output power is generated for typical wind conditions.

[0004] Nevertheless, the structure and components of a wind turbine are subject to a plurality of loads having different magnitudes and frequencies during operation. In order to ensure the structural integrity and proper operation of the wind turbine, it is necessary to maintain such loads within a certain range that defines the operating limit values of the wind turbine (for example, the maximum rotational speed or the maximum electromagnetic torque). Therefore, one of the purposes of the control system is to control the actuator so that the wind turbine is maintained within the operating limit values.

[0005] Therefore, the controller or control system is the most important part of the wind turbine system. The ability to obtain energy from the wind and the cost of the wind turbine greatly depend on the ability of the controller to reduce the loads on the components of the wind turbine (such as, but not limited to, blades, hub, tower, bearing structure, etc.). Various control strategies have been used for the control of wind turbines. These strategies include classical PID-type feedback controllers, which are widely used and have achieved reliable operation. However, wind turbines operate in a varying environment and are becoming increasingly complex systems, and advanced control strategies that can control systems with multiple control variables have also been proposed. Among advanced modern control strategies, model predictive control (MPC) is particularly suitable for the control of wind turbines. This is because MPC can handle multivariable control problems while considering process constraints. Therefore, the MPC controller can also consider structural loads (such as the moment at the blade root section, the moment at the base of the tower) to maintain the stability of the system while optimizing the output power, and manage the constraints on the inputs (such as the pitch angle of the blade or the torque of the generator) and the constraints on the output or operating performance (such as the rotational speed, generated power).

[0006] To determine the control action, the MPC uses a mathematical model of the wind turbine to predict the future behavior (i.e., output variables) of the wind turbine over a specific time horizon after considering the current state of the wind turbine. The predicted behavior of the wind turbine depends not only on the current state of the wind turbine but also on changes in controllable variables and environmental conditions during the prediction period. The MPC controller provides an optimal control action as an output to the actuators of the wind turbine, thereby ensuring that the wind turbine reliably follows the desired reference values. For this purpose, the MPC controller includes an optimizer that solves an online optimization problem at each control step. The optimization problem involves minimizing a cost function J over the prediction period under several constraints. This cost function penalizes the deviation of the output variables (also called dependent or control variables) from the reference values (such as output power and speed reference). In addition, the cost function also penalizes excessive movement or changes in the input variables (also called independent or controllable variables). As those skilled in the art will recognize, both the cost function and the constraints can be defined based on the specific objectives and / or requirements of the control problem.

[0007] To enhance the competitiveness of the wind turbine design, it is important for the controller to handle the loads on the turbine components. Especially when using an MPC controller, the loads can be treated as constraints in the optimization problem. Therefore, the constraints are introduced to consider the limits of the system. However, it is not practical to evaluate the strength of each component under each operating condition online because it requires a large number of detailed calculations, i.e., significant computational power. For this reason, the load evaluation during the design of the wind turbine is usually simplified by selecting a specific limited number of components and component positions (e.g., blade root section). Furthermore, a limited number of secondary load parameters for those specific positions and components (e.g., blade flap moment at the blade root section) are also selected. This known simplified approach attempts to capture the full load envelope.

[0008] However, this known method has at least two disadvantages. One disadvantage is that when a specific load condition is reached and that specific load condition exceeds the strength limit value of a certain component, but the component is not considered as a specific component and the secondary load parameters are not considered either, safety problems may occur. In such a case, the controller may not be able to take corresponding measures to reduce the load so that the wind turbine is maintained within the operation limit value. The second disadvantage caused by this known method relates to using an excessive (i.e., too large) safety margin to prevent the aforementioned problems. The safety margin can be used, for example, to address the uncertainty of the strength limit value resulting from not considering the most critical components and the most critical load conditions in a simplified method. Although this method can indeed reduce safety concerns, selecting a high safety margin may make the design of the wind turbine overly conservative during the transition, and as a result, the cost may become too high.

[0009] The present disclosure aims to provide an improved method for controlling a wind turbine using a predictive controller so that at least some of the aforementioned disadvantages are reduced. SUMMARY OF THE INVENTION

[0010] In one aspect of the present disclosure, a method (100) for controlling a wind turbine (10) having a plurality of wind turbine actuators (364) is provided. The method has the following steps. First, the method includes receiving operating data (366) of the wind turbine (10). Based on the received operating data (366), the operating state of the wind turbine (10) is estimated. The method further includes predicting a possible operating state of the wind turbine (10) according to the operation of the wind turbine actuator (364) over a finite time using a control model (370). The control model (370) includes a predictive aeroelastic model (371) for determining a primary load (375) based on the received operating data (366). The control model (370) further includes a strength calculation module (372) for calculating one or more secondary load parameters (374) based on the primary load (375), and constraints are defined for the secondary load parameters (374). The method includes optimizing a cost function over a finite optimization time according to the constraints to determine an optimal trajectory (391) including commands for the wind turbine actuator (364). Then, the method includes controlling the wind turbine actuator (364) using the first command (367) of the determined optimal trajectory.

[0011] According to this aspect of the present disclosure, a method for improving the control of a wind turbine is provided. The output power is optimized without exceeding the operating limit values of the wind turbine. By using the control model of the wind turbine, the predictive ability of the model is utilized to optimize future control actions.

[0012] This method optimizes the cost function over an optimization time such that the optimal operation of the wind turbine actuator is defined. In an example of the present disclosure, the optimization time coincides with the time used for predicting the possible operating states of the wind turbine. In other examples, different times can be used for optimization and prediction. The cost function can combine and achieve multiple objectives. More specifically, the optimization of the cost function follows constraints representing different system limit values.

[0013] This model includes a strength calculation module that can calculate secondary load parameters based at least in part on the primary loads determined by the predictive aeroelastic model. On the other hand, the secondary load parameters correspond to the most relevant loads that truly determine the limitations of the wind turbine. On the other hand, the primary loads are provided by the predictive aeroelastic model, and the primary loads correspond to the basic loads obtained from simplified calculations. This basic load may be given, for example, by the bending moment at a selected interface between different selected components of the wind turbine and / or the force at a selected position of a specific component. In some cases, the basic load may be expressed as the deflection of a specific component.

[0014] Constraints are defined for the secondary load parameters. The calculated values of the secondary load parameters and the corresponding constraints are used in combination with the cost function to constitute an optimization problem solved by the MPC controller. At least some of the multiple constraints used in the optimization process are related to the secondary load parameters calculated by the strength calculation module. As a result, an optimal command for the wind turbine actuator is provided, taking into account the most relevant loads that affect the operation of the wind turbine.

[0015] Unlike the control methods known in the art, a representation that is truly relevant to the load in real time and is sufficiently close to it is used. The relevant load is characterized by detailed secondary load parameters calculated by a strength calculation module. In fact, in this control method, instead of using a limited set of specific loads at specific positions, the true limiting conditions are considered. The set of specific loads at specific positions can correspond to a set of basic loads (such as basic loads provided by a simplified predictive aerodynamic elastic model). Those basic loads can, for example, conform to examples of design loads derived from relevant standards in this field. However, these basic loads may not represent the true system limitations. For this reason, in a control system based on such basic loads, typically, a very conservative safety margin needs to be used. According to the present disclosure, since secondary load parameters indicating the true limiting conditions are calculated and used by the controller, more accurate safety information is used in the control of wind turbines. As a result, the handling of loads is improved, and it is possible to prevent the design of wind turbines from becoming too conservative.

[0016] In another aspect of the present disclosure, a controller for a wind turbine is provided. The control system is characterized in that it is configured to execute the method according to the foregoing aspect.

[0017] According to other aspects of the present disclosure, a controller or control system can be disposed in a wind turbine. The control system includes hardware necessary to execute a method for controlling the wind turbine. More specifically, the control system includes hardware that stores a model of the wind turbine including a strength calculation module, and an optimization algorithm including a cost function and constraints. In this way, the control system can optimally control the wind turbine by being able to execute the aforementioned method. Thus, the controller is configured such that load handling is improved and violations of design limitations of wind turbine components during operation are avoided. By improving load handling, true limiting factors are considered and the annual energy production (AEP) can be increased. In fact, by using more accurate constraints, the controller does not need to add conservative limit values considering uncertainties. As a result, the wind turbine can obtain more energy than other methods, for example, by using a temporarily low pitch angle or a large generator torque. Further, a wind turbine incorporating a control system according to this aspect of the present disclosure can avoid the costs incurred when applying overly conservative limit values to different components and can be designed in a more competitive way.

[0018] Throughout the present disclosure, the term secondary load parameter is interpreted to refer to a detailed magnitude obtained by a strength calculation module that represents the true limiting loads to which one or more wind turbine components are subjected. Thus, by way of non-limiting example, a secondary load parameter represents a force or moment to which a wind turbine component is subjected, although a secondary load parameter may also refer to a derived parameter (such as stress or strain) of a wind turbine component. A secondary load parameter can also be defined in terms of other secondary load parameters after they are combined. A secondary load parameter is determined by a strength calculation module after considering the primary loads. Throughout the present disclosure, the primary loads represent the basic loads that can be obtained from a simplified load calculation in a predicted aeroelastic control model. Examples of primary loads can include, for example, a bending moment at the interface of a selected component of a wind turbine, or a force at a selected location of a particular component. Specifically, examples of such primary loads can include the flapwise and edgewise moments at the blade root, or the nodding moment at the tower base. A primary load can also be expressed as the deflection of a particular component.

Brief Description of the Drawings

[0019] Hereinafter, non-limiting examples of the present disclosure will be described with reference to the drawings.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Modes for Carrying Out the Invention

[0020] Multiple embodiments are referred to in detail. One or more examples of the multiple embodiments are shown in the drawings. Each example is provided for illustrative purposes only and not for limitation. In fact, it is obvious to those skilled in the art that various modifications and variations can be made in the present disclosure. For example, features illustrated or described as part of one example can be used in combination with another example to realize yet another example. Therefore, the present disclosure is intended to cover modifications and variations that fall within the scope of the claims and their equivalents.

[0021] FIG. 1 is a perspective view of an example of a wind turbine 10. In this example, the wind turbine 10 is a horizontal axis wind turbine. Alternatively, the wind turbine 10 may be a vertical axis wind turbine. In an embodiment, the wind turbine 10 includes a tower 15 extending from a support system 14 on the ground 12, a nacelle 16 attached to the tower 15, and a rotor 18 coupled to the nacelle 16. The rotor 18 includes a rotatable hub 20 and at least one rotor blade 22 coupled to the hub 20 and extending outwardly from the hub 20. In an embodiment, the rotor 18 has three rotor blades 22. Alternatively, the rotor 18 may include more or fewer than three rotor blades 22. The tower 15 can be made of tubular steel such that a cavity (not shown in FIG. 1) is defined between the support system 14 and the nacelle 16. Alternatively, the tower 15 is a suitable type of tower having a suitable height. According to an alternative, the tower can be a hybrid tower including a concrete portion and a steel pipe portion. Also, the tower can be a partial lattice tower or a full lattice tower.

[0022] The rotor blades 22 are spaced apart about the hub 20 so as to enable the rotor 18 to be rotated to utilize the kinetic energy from the wind as mechanical energy available and then converted into electrical energy. The rotor blades 22 are fitted to the hub 20 by coupling the blade root portion 24 to the hub 20 in a plurality of load transfer regions 26. The load transfer regions 26 can have a hub load transfer region and a blade load transfer region (both not shown in FIG. 1). The loads induced on the rotor blades 22 are transmitted to the hub 20 through the load transfer regions 26.

[0023] In a plurality of examples, the rotor blade 22 can have a length ranging from about 15 meters (m) to about 90 m or more. The rotor blade 22 can have an appropriate length to enable the wind turbine 10 to function as described herein. For example, non-limiting examples of blade lengths include lengths less than 20 m, 37 m, 48.7 m, 50.2 m, 52.2 m, or greater than 91 m. When the wind hits the rotor blade 22 from the wind direction 28, the rotor 18 rotates about the rotor shaft 30. When the rotor blade 22 rotates and is subjected to centrifugal force, the rotor blade 22 is also subjected to various forces and moments. Therefore, the rotor blade 22 deflects and / or rotates from the neutral position (or non-deflected position) to the deflected position.

[0024] Furthermore, the pitch angle of the rotor blade 22, that is, the angle that determines the orientation of the rotor blade 22 with respect to the wind direction, is changed by the pitch system 32, and by relatively adjusting the angular position of at least one rotor blade 22 with respect to the wind vector, the load and output power of the wind turbine 10 can be controlled. The pitch axis 34 of the rotor blade 22 is shown. During operation of the wind turbine 10, the pitch system 32 can change the pitch angle of the rotor blade 22 such that, in particular, the angle of attack of (a part of) the rotor blade is reduced, thereby promoting a decrease in the rotational speed and / or promoting a stall of the rotor 18.

[0025] In this embodiment, the blade pitch of each rotor blade 22 is individually controlled by the wind turbine controller 36 or by the pitch control system 80. Alternatively, the blade pitch of all the rotor blades 22 may be controlled simultaneously by the control system.

[0026] Furthermore, in the embodiment, when the wind direction 28 changes, the nacelle 16 rotates about the longitudinal axis of the tower, that is, the yaw axis 38, and the rotor blade 22 can be positioned relative to the wind direction 28.

[0027] In this embodiment, the wind turbine controller 36 is shown as being centrally located within the nacelle 16. However, the wind turbine controller 36 may be a system that is distributed throughout the wind turbine 10, the support system 14, within the wind farm, and / or at a remote control center. The wind turbine controller 36 can include a processor 40 configured to execute some of the methods and / or steps described herein. Further, many of the other components described herein include a processor.

[0028] As used herein, the term "processor" is not limited to integrated circuits referred to as computers in the art, but broadly refers to controllers, microcontrollers, microcomputers, programmable logic controllers (PLCs), application specific integrated circuits, and other programmable circuits, and these terms are used interchangeably herein. It should be understood that the processor and / or control system can also include a memory, an input channel, and / or an output channel.

[0029] The control system 36 can also include a memory (e.g., one or more memory devices). The memory can include memory elements, and examples of memory elements include computer-readable media (e.g., random access memory (RAM)), computer-readable non-volatile media (e.g., flash memory), floppy disks, compact disc read-only memory (CD-ROM), magneto-optical disks (MOD), digital versatile disks (DVD), and / or other suitable memory elements, but are not limited thereto. Such memory devices can generally store appropriate computer-readable instructions that, when executed by the processor 40, configure the controller 36 to execute various steps disclosed herein or trigger the execution of various steps. The memory can also be configured to store, for example, data obtained by measurement and / or data obtained by calculation.

[0030] FIG. 2 is an enlarged cross-sectional view of a portion of the wind turbine 10. In this example, the wind turbine 10 includes a nacelle 16 and a rotor 18 rotatably coupled to the nacelle 16. More specifically, the hub 20 of the rotor 18 is rotatably coupled to a generator 42 disposed within the nacelle 16 by a main shaft 44, a gearbox 46, a high-speed shaft 48, and a coupling 50. In an embodiment, the main shaft 44 is disposed at least partially coaxially with the longitudinal axis (not shown) of the nacelle 16. When the main shaft 44 rotates, the gearbox 46 is driven, and subsequently, the gearbox 46 drives the high-speed shaft 48 by converting the relatively slow rotational movement of the rotor 18 and the main shaft 44 into the relatively fast rotational movement of the high-speed shaft 48. The high-speed shaft 48 is connected to a generator 42 for generating electrical energy by a coupling 50. Further, in order to convert the electrical energy generated by the generator 42 having a voltage of 400V to 1000V into electrical energy having a medium voltage (e.g., 10KV to 35KV), a transformer 90 and / or appropriate electronic devices, switches, and / or inverters can be disposed in the nacelle 16. The electrical energy is transmitted from the nacelle 16 to the tower 15 through a power cable.

[0031] The gearbox 46, the generator 42, and the transformer 90 are supported by a main support structure frame of the nacelle 16, and the main support structure frame can be embodied as a main frame 52 optionally. The gearbox 46 can include a gearbox housing connected to the main frame 52 by one or more torque arms 55. In an embodiment, the nacelle 16 also includes a main front support bearing 60 and a main rear support bearing 62. In particular, the generator 42 can be attached to the main frame 52 by decoupling support means 54 to prevent the vibration of the generator 42 from being transmitted to the main frame 52 and thereby becoming a noise emission source.

[0032] Optionally, the main frame 52 is configured to carry the overall loads caused by the weights of the components of the rotor 18 and the nacelle 16 as well as by the wind loads and the rotational loads, and further to transmit these loads to the tower 15 of the wind turbine 10. The rotor shaft 44, the generator 42, the gearbox 46, the high-speed shaft 48, the coupling 50, and the fastening devices, support devices, and / or fixing devices (such as, but not limited to, the support frame 52, the front support bearing 60, and the rear support bearing 62) may be referred to as a drive train 64.

[0033] In some embodiments, the wind turbine may be a direct-drive wind turbine without a gearbox 46. The generator 42 operates at the same rotational speed as the rotor 18 of the direct-drive wind turbine. Thus, a direct-drive wind turbine generally has a much larger diameter than the generator used in a wind turbine having a gearbox 46 in order to supply generally the same amount of power as a wind turbine having a gearbox.

[0034] The nacelle 16 can also include a yaw system, the yaw system including a yaw bearing (not visible in FIG. 2) having two bearing elements, the two bearing elements being configured to rotate relative to the other bearing element. The tower 15 is coupled to one bearing element and the bedplate or support frame 52 of the nacelle 16 is coupled to the other bearing element.

[0035] The yaw system includes a ring gear 31 and a yaw drive mechanism 56, and the yaw system can be used to rotate the nacelle 16 about the longitudinal axis of the tower (i.e., the yaw axis 38 that controls the position of the rotor blades 22 with respect to the wind direction 28), thereby also rotating the rotor 18.

[0036] The yaw drive mechanism 56 can include a plurality of yaw drive units 35, and the plurality of yaw drive units 35 have a motor 33, a gearbox 37, and a pinion 39 that meshes with an annular gear 31 to rotate one bearing element relative to the other bearing element. The annular gear 31 can include a plurality of teeth that mesh with the teeth of the pinion 39. In the example of FIG. 2, the yaw drive unit 35 and the annular gear 31 are disposed outside the outer diameter of the tower. The teeth of the annular gear are directed outward, but in other examples, the annular gear and the yaw drive unit may be disposed inside the tower.

[0037] In some examples, one of the plurality of yaw drive units can be designated as the "master", and the other drive units can be designated as "slaves" that can follow the instructions of the master or adapt their operation to match the master drive unit.

[0038] The wind turbine controller 36 can be communicatively coupled to the yaw drive mechanism 56 of the wind turbine 10 to control and / or change the yaw direction of the nacelle 16 relative to the wind direction 28. When the wind direction 28 changes, the wind turbine controller 36 can control the yaw angle of the nacelle 16 with respect to the longitudinal axis or yaw axis 38 of the tower to position the rotor blades 22, and thus the rotor 18, relative to the wind direction 28, thereby controlling the load acting on the wind turbine 10. For example, the wind turbine controller 36 can be configured to transmit a control signal or command to the yaw drive mechanism 56 of the wind turbine 10 through a yaw controller or by direct drive, and the nacelle 16 can rotate about the longitudinal axis or yaw axis 38 of the tower via a yaw bearing.

[0039] To properly position the nacelle 16 relative to the wind direction 28, the nacelle 16 can also include at least one weather measurement system that includes a wind vane and an anemometer. The weather measurement system 58 can provide information including the wind direction 28 and / or wind speed to the wind turbine controller 36.

[0040] In this embodiment, at least a part of the pitch system 32 is arranged as a pitch assembly 66 within the hub 20. The pitch assembly 66 includes one or more pitch drive systems 68 and at least one sensor 70. Each pitch drive system 68 is coupled to a respective rotor blade 22 (shown in FIG. 1) to adjust the pitch angle of the rotor blade 22 along the pitch axis 34. Only one of the three pitch drive systems 68 is shown in FIG. 2.

[0041] In the embodiment, the pitch assembly 66 includes at least one pitch bearing 72 coupled to the hub 20 and a respective rotor blade 22 (shown in FIG. 1) to rotate the respective rotor blade 22 about the pitch axis 34. The pitch drive system 68 includes a pitch drive motor 74, a pitch drive gearbox 76, and a pitch drive pinion 78. The pitch drive motor 74 is coupled to the pitch drive gearbox 76 to apply mechanical force to the pitch drive gearbox 76. The pitch drive gearbox 76 is coupled to the pitch drive pinion 78, and the pitch drive pinion 78 is rotated by the pitch drive gearbox 76. The pitch bearing 72 is coupled to the pitch drive pinion 78, and the pitch bearing 72 rotates when the pitch drive pinion 78 rotates.

[0042] The pitch drive system 68 is coupled to the wind turbine controller 36 and adjusts the pitch angle of the rotor blades 22 when receiving one or more signals from the wind turbine controller 36. In this embodiment, the pitch drive motor 74 is any suitable motor driven by an electric power and / or hydraulic system such that the pitch assembly 66 can function as described herein. Alternatively, the pitch assembly 66 can include any suitable structure, configuration, arrangement, and / or components (such as, but not limited to, hydraulic cylinders, springs, and / or servo mechanisms). In a particular embodiment, the pitch drive motor 74 is driven by energy extracted from the rotational inertia of the hub 20 and / or an energy storage source (not shown) that supplies energy to components of the wind turbine 10.

[0043] The pitch assembly 66 also includes one or more pitch control systems 80 for controlling the pitch drive system 68 in accordance with control signals from the wind turbine controller 36 in certain prioritized situations and / or while the rotor 18 is over-rotating. In an embodiment, the pitch assembly 66 includes at least one pitch control system 80 communicatively coupled to each pitch drive system 68, the at least one pitch control system 80 being capable of controlling the pitch drive system 68 independently of the wind turbine controller 36. In an embodiment, the pitch control system 80 is coupled to the pitch drive system 68 and the sensor 70. During normal operation of the wind turbine 10, the wind turbine controller 36 can control the pitch drive system 68 such that the pitch angle of the rotor blades 22 is adjusted.

[0044] According to one embodiment, the generator 84 has, for example, a battery and an electrical capacitor, and the generator 84 is disposed on or within the hub 20 and is coupled to the sensor 70, the pitch control system 80, and the pitch drive system 68, serving as a power source for these components. In this embodiment, the generator 84 serves as a continuous power source for the pitch assembly 66 during the operation of the wind turbine 10. In an alternative example, the generator 84 supplies power to the pitch assembly 66 only while an event of power loss occurs in the wind turbine 10. The power loss event includes a loss or dip in the power grid, a failure of the electrical system of the wind turbine 10, and / or a failure of the wind turbine controller 36. While the power loss event is occurring, the generator 84 operates to supply power to the pitch assembly 66 so that the pitch assembly 66 can operate.

[0045] In this example, each of the pitch drive system 68, the sensor 70, the pitch control system 80, the cable, and the generator 84 is disposed within a cavity 86 defined by the inner surface 88 of the hub 20. In an alternative embodiment, the components are disposed in a portion corresponding to the outer surface of the hub 20 and can be directly or indirectly coupled to the outer surface.

[0046] FIG. 3 is a diagram showing a system having a wind turbine 10, a controller 360, and a wind turbine actuator system 364. The wind turbine actuator system 364 includes a plurality of actuators (such as the blade pitch system 32, yaw drive mechanism 56, or an electronic converter enabling torque control described above). Although depicted separately in FIG. 3 for purposes of illustration and explanation, those actuators are physically disposed within the wind turbine 10 as would be well understood by those skilled in the art. The controller 360 of the present embodiment has an estimator 361 and a MPC (Model Predictive Control) module 362. The controller 360 controls the wind turbine 10 while maintaining system limitations such that the wind turbine 10 operates in accordance with one or more received reference signals 365 (such as power output). For this purpose, the controller 360 generates instructions (or instruction 367) for controlling one or more wind turbine actuators of the wind turbine actuator system 364. A plurality of wind turbine actuators may be present in the wind turbine 10. Specifically, in an embodiment of the present disclosure, the actuator instructions can include at least one of a blade pitch angle and a generator torque and can be transmitted to the corresponding blade pitch drive system 68 and power converter.

[0047] To control the wind turbine 10, the controller 360 can receive operating data 366 from the wind turbine 10. The operating data 366 can be used to determine the operating state of the wind turbine 10. In an example of the present disclosure, the operating data can include values such as a blade pitch angle, output power, electrical torque, rotational speed of the wind turbine rotor, rotational speed of the generator, tower top position, and the like. The operating data 366 can be obtained by direct measurement by sensors. However, in an embodiment of the present disclosure, the operating data 366 may be obtained by a virtual sensor. Thus, in an embodiment of the present disclosure, direct measurement may not be necessary. A virtual sensing system including an appropriate model may be used to calculate the value of an operating variable when a measured value of the operating variable cannot be directly obtained.

[0048] In an embodiment of the present disclosure, sensor measurements can be supplied to a signal processing module (not shown in FIG. 3), and the signal processing module can filter and adjust the sensor measurements. And an estimator 361 can be provided. The estimator 361 can include a filter equation (e.g., an extended Kalman filter) that can be used to calculate the most likely operating state of the wind turbine 10 based on the received operating data 366. The operating state can be determined dynamically at each control step.

[0049] As also shown in FIG. 3, the determined operating state can be used by the MPC module 362. The MPC module 362 defines an actuator command based on the current operating state and control requirements of the wind turbine 10. For this purpose, the MPC module 362 uses a model of the wind turbine 10. Details of the MPC module 362 will be described with reference to FIGS. 5 and 6.

[0050] FIG. 4 shows a flowchart of an example of a method 100 for controlling a wind turbine 10 having a plurality of wind turbine actuators 364. The method 100 includes steps 110 to 150. Step 110 includes receiving operating data 366 of the wind turbine 10. Next, step 120 includes estimating the operating state of the wind turbine 10 based on the received operating data 366. Step 130 includes predicting a possible operating state of the wind turbine according to the operation of the wind turbine actuator 364 over a finite time using a control model (370 in FIGS. 5 and 6). The control model 370 includes a predictive aeroelastic control model (371 in FIGS. 5 and 6) for determining a primary load (375 in FIGS. 5 and 6) based on the received operating data 366. The control model 370 further includes a strength calculation module (372 in FIGS. 5 and 6) that calculates one or more secondary load parameters (374 in FIGS. 5 and 6) based on the primary load 375 determined by the predictive aeroelastic control model 371, and constraints are defined for the secondary load parameters 374. In step 140, an optimal trajectory including commands for the wind turbine actuator is determined by optimizing a cost function over a finite time according to the constraints. Finally, step 150 includes controlling the wind turbine actuator 364 using the first command of the determined optimal trajectory.

[0051] In an embodiment of the present disclosure, the method can include repeating the above steps 110 to 150 at successive control intervals. More specifically, steps 110 to 150 can be executed at each control interval. In this way, the optimal control of the wind turbine 10 can be executed in real time so as to preferably respond and adapt to changing environmental conditions and / or changing operating conditions of the wind turbine 10.

[0052] In embodiments of the present disclosure, the method can include defining constraints not only for secondary load parameters but also for one or more primary loads. As a result, in these examples, optimizing the cost function over a finite optimization time can also comply with the constraints defined for the primary loads. Thus, according to these examples, the behavior of the wind turbine can be more adequately captured using a combination of constraints related to the primary loads and the secondary load parameters.

[0053] Furthermore, in examples of the present disclosure, the secondary load parameters may be related to extreme loads. In other examples, fatigue loads may also be considered in the definition of the secondary load parameters.

[0054] The operating state of the wind turbine 10 is defined in step 120 of method 100 and can be defined by a plurality of variables related to the operating parameters of the wind turbine 10. The variables can be represented as a vector. As described with reference to FIG. 3, the operating state can be obtained after considering direct sensor measurements or virtual sensor information. Further, an estimator 361 is used to determine the current operating state, and thus uncertainties related to, for example, noise can be reduced. Different variables (i.e., different operating parameters) can be selected to define the operating state of the wind turbine 10. In an example of the present disclosure, the operating state of the wind turbine 10 can be defined as follows.

Number

[0055] Here, α is the blade pitch angle, ω is the rotational speed of the wind turbine rotor, T is the electromagnetic torque of the generator, b is the blade deflection, and t is the tower deflection. The time derivative of each of these parameters can also be included in the definition of the operating state.

[0056] The optimal trajectory for controlling the wind turbine actuator 364 includes a time series of commands that can be described as a vector including the commands for the wind turbine actuator 364. According to the present disclosure, the first value of the trajectory is directly given to the wind turbine actuator 364. Similar to the definition of the operating state of the wind turbine 10, in the embodiments of the present disclosure, different trajectories can be assumed according to the specific wind turbine actuators 364 considered (and the specific parameters of those actuators). Specifically, the trajectory provides a list of controllable variables or independent variables used to dynamically control the behavior of the wind turbine 10. In one embodiment of the present disclosure, the optimal trajectory can include the following commands for the blade pitch angle and the generator torque.

Number

[0057] Here, α i represents the reference for the pitch angle of blade “i” of the wind turbine 10, and T r corresponds to the reference for the electric torque. In another embodiment, the optimal trajectory can include the command for the rate of change of the blade pitch angle and the command for the generator torque.

[0058] Step 130 of the method includes predicting the possible operating states of the wind turbine 10 according to the operation of the wind turbine actuator 364 using the model of the wind turbine 10. In other words, the model is used to predict the future trajectory, i.e., the time variation, of the operating state of the wind turbine 10. This operating state can be defined according to the output variables or dependent variables as shown above. The trajectory depends on the current operating state (determined in step 120) and the operations on the controllable variables or independent variables (e.g., blade pitch or torque). Apart from the controllable variables, the influence of the uncontrollable independent variables can also be considered by the model in the form of disturbances. The determination of the possible trajectory includes the calculation of specific secondary load parameters 374 indicating the predicted loads of the selected components of the wind turbine 10.

[0059] The control model 370 includes a predictive aeroelastic model 371 used to determine the primary load 375. The control model 370 further includes another model, namely a strength calculation module 372, which is used to calculate the truly important secondary load parameter 374 for determining the behavior of the wind turbine 10. By adding the strength calculation module 372, the secondary load parameter 374 beyond the range usually provided by the predictive aeroelastic model 371 can be calculated and used in the optimization problem. The strength calculation module 372 corresponds to the code used to calculate the secondary load parameter 374. The strength calculation module 372 may be implemented as a separate block or as a specific block, or may be incorporated into other parts of the code. In other words, although the strength calculation module 372 is shown as an independent module for illustration purposes in FIGS. 5 and 6, the function of the strength calculation module 372 may be integrated with other parts of the code.

[0060] From among the predicted trajectories for the operating state of the wind turbine 10, an optimal trajectory is obtained in step 140. For this purpose, a cost function is minimized according to a number of constraints. At least some of the multiple constraints used in the optimization problem are related to the calculated secondary load parameter 374.

[0061] An example of a general cost function is shown below.

Equation

[0062] In the cost function of this example, for the operating state of wind turbine 10, N dependent variables or output variables x i are defined, and for the actuators of the wind turbine, M independent variables or manipulation variables u i are defined.

[0063] The first term of the cost function J corresponds to the output variables or dependent variables. The first term ensures that the resulting operating state of wind turbine 10 closely follows the desired criteria for the output variables. The second term corresponds to the variation of the controllable variables or independent variables (i.e., the commands that control the actuators of the wind turbine). This term is used to avoid excessive movement of the actuators. The individual weights W i are used to prioritize the performance goals of the controller by adjusting the tuning weights of the cost function. Generally, increasing the weight of the output variables results in the performance of actively following the reference values, and increasing the weight of the controllable variables or independent variables can smooth the control operation and improve robustness.

[0064] In an example of the present disclosure, the cost function can be defined as follows.

Equation

[0065] Here, the first two terms correspond to the tracking of power P and rotor speed ω. In other words, the first two terms of this cost function are such that the output variables are the respective reference values P ref and ω refTo ensure exact compliance. The remaining two terms relate to the blade pitch (q) and torque (T), which are independent variables or controllable variables, or specifically to the rate of change of pitch and the rate of change of torque. These two terms are included to characterize the operation of the controllable variables and to impose a penalty. Finally, to account for the relative importance of each variable, an individual weight (W i ) can be assigned to each term.

[0066] As described above, since different output variables can be used to characterize the operating state of the wind turbine 10 and different controllable variables can be used for different wind turbine actuators, this cost function is shown as just an example.

[0067] In an embodiment of the present disclosure, the secondary load parameter can be included in the cost function. Specifically, the secondary load parameter can be included together with a corresponding weight indicating the relative importance of the secondary load parameter. Further, in a variant of the embodiment, the constraints defined for the secondary load parameter can include at least an upper limit and a slack variable, and the slack variable is penalized in the cost function.

[0068] According to this embodiment, the constraints related to the secondary load parameter can be included as soft constraints in the optimization problem. In this way, if the constraints set for the secondary load parameter are included in the problem as hard constraints, the constraints set for the secondary load parameter may cause an unsolvable optimization problem. To prevent such a situation, the constraints can be included as soft constraints using a slack variable. The corresponding constraints can be realized by defining an inequality constraint including the slack variable in the problem and adding a term that penalizes the slack variable to the cost function. The size of the slack variable can correspond to the size of the related constraint violation. By adding the slack variable to the cost function, the optimizer of the optimization problem can search for an optimal trajectory while keeping the slack (i.e., the constraint violation) as small as possible.

[0069] Therefore, in a general form, the constraint of the secondary load parameter LP can be defined as follows by inequality constraints.

Equation

[0070] Here, UB represents the upper limit of each secondary load parameter LP, and ε is the slack related to the secondary load parameter. By adding a slack variable to the aforementioned cost function, a modified general cost function as shown by the following equation can be obtained.

Equation

[0071] Here, W LP represents the weight of the related secondary load parameter LP, which indicates the relative importance of the corresponding secondary load parameter. Also, in this example, the slack is introduced in the form of a quadratic function. However, this is just an example, and other functions f(ε) can also be used as long as it is a monotonically increasing function of ε.

[0072] While defining the constraint of the secondary load parameter as a soft constraint, the values of the upper limit UB and the slack ε can be adjusted respectively so that the controller realizes the desired operation.

[0073] In other embodiments of the present disclosure, the secondary load parameter can be added to the cost function as one of the tracking members of the cost function. In other words, a specific value can be defined for a specific secondary load parameter, and terms equivalent to those listed above for the output power or the rotor speed can be included in the cost function of the secondary load parameter.

[0074] In an example of the present disclosure, the constraints defined for the secondary load parameters are considered to be based on the limiting values of the materials of one or more wind turbine components. In fact, the limiting values of the materials of the wind turbine components can be used as constraints in the optimization problem. The reason is that they must not exceed the limiting values of the materials. As shown below, the secondary load parameters indicating the limiting values of the materials can be given in different magnitudes (forces, moments, stresses, strains, etc.).

[0075] Furthermore, as already shown in connection with FIGS. 1 and 2, the wind turbine 10 is a complex system having a very large number of components. Each of these components is manufactured from different materials and may be subject to different operating loads. As a result, the limiting values of the materials will be represented in different forms for different parts. In an example of the present disclosure, examples of the components include blades, towers, hubs, bed plates, foundations, and the like.

[0076] In one embodiment of the present disclosure, the secondary load parameters can be predetermined. More specifically, the secondary load parameters can be predetermined based on a finite element method analysis of selected components of the wind turbine 10. According to this embodiment, the determination of which secondary load parameters to use as constraints in the optimization problem can be made in advance. Based on that determination, the strength calculation module 732 can be programmed, and the strength calculation module 732 can include equations for calculating the values of the selected secondary load parameters. The determination of the values of the secondary load parameters can include calculations and linearization of the calculations. Linearization enables online calculations, and the secondary load parameters can be used efficiently by an optimization algorithm during the real-time operation of the wind turbine 10. This is especially true for particularly complex secondary load parameters.

[0077] During the design phase of the wind turbine 10, detailed simulations can be performed for each of the most relevant components. Specifically, the simulations can be performed using a finite element method (FEM) model. In this way, the secondary load parameters that define the true mechanical limit values of each component of the wind turbine 10 can be identified.

[0078] As already shown, different secondary load parameters can be used depending on the requirements of the wind turbine 10 and the operating conditions of the wind turbine. In the examples of the present disclosure, a given secondary load parameter can include at least one of force, moment, stress, strain, or buckling load. More specifically, the secondary load parameter may include the stress in a wind turbine component derived from a stress tensor, or the strain of a wind turbine component derived from a strain tensor.

[0079] The design limit values can be defined by secondary load parameters that represent only the force or moment at a specific location of a wind turbine component. In such a case, the soft constraint can be defined as follows. Force or moment x ≦ UB + ε

[0080] Here, "force or moment x " simply represents the value of the selected force or moment for the component "x" of the wind turbine 10. However, in other embodiments of the present disclosure, by assuming more detailed secondary load parameters, the limiting loads of specific components can be grasped. In such a case, a function of multiple components of force or moment can be used. For this purpose, the three-dimensional force (F x , F y , F z ) and the three-dimensional moment (M x , M y , M z ) can be represented by the resultant force R.

Equation

[0081] Next, the critical secondary load parameter can be obtained as a function of the resultant force R. Further, the secondary load parameter can be derived using a plurality of resultant forces at different positions of the component. As a result, the corresponding constraints are generally defined as follows.

Equation

[0082] Here, LP_i represents a specific secondary load parameter of the wind turbine component among p different secondary load parameters, and R1...R N represents the resultant force and moment at N different positions. Finally, UB i represents the upper limit value of the corresponding secondary load parameter LP_i.

[0083] In the example of the present disclosure, the secondary load parameter derived from the resultant force R can be given by stress or strain. Accordingly, the wind turbine component is subjected to a normal force, a shear force, a bending moment, and a torsional moment. The load state of a part of the component can be calculated from the resultant force R using standard techniques in elastic engineering. From these calculations, a stress (σ) tensor or a strain (ε) tensor can be derived.

Equation

[0084] These tensors represent the structural state of the component. For the safety of the component under complex loading conditions, it may be required that the material of the component does not exceed the critical stress or critical deformation. Such a constraint is expressed as follows in the optimization problem.

Equation

[0085] Here, F represents a secondary load parameter that is a function of the stress, strain, or other load factor of the component at a specific geometric location of the component. Using several scales, the limit value F cr can be expressed, and the limit value F cr depends on the limit state of the given material. These selectable load factors include those where F cr corresponds to the maximum principal stress, the maximum shear stress, the maximum principal strain, the maximum strain energy, or the maximum shear strain energy. By this general expression, flexibility can be enhanced when dealing with the limiting conditions corresponding to different types of materials for different components of the wind turbine. In the examples of the present disclosure, composite materials can be considered. This applies when dealing with the limit value of blade delamination in the blade composite. In such a case, the secondary load parameter F can correspond to the complementary strain energy.

[0086] FIG. 5 shows a more detailed view of an example of a controller 360 for implementing the method as described above. As shown in the figure, the control system 360 can be implemented using a model predictive control (MPC) module 362. As shown previously, an estimator 361 can be used to characterize the operating state of the wind turbine 10. The operating state of the wind turbine 10 is supplied to a control model 370 of the MPC module 362. A predictive aeroelastic control model 371 is used to determine a specific primary load 375 for the selected components and locations of the wind turbine 10. The primary load 375 can, in some cases, be represented as the deflection of a specific component of the wind turbine 10. A strength calculation module 372 uses the primary load 375 from the predictive aeroelastic control model 371 to calculate the value of a secondary load parameter 374. The secondary load parameter 374 can be associated with the corresponding constraints for the optimization problem, and this constraint can be defined in the corresponding constraint module 373.

[0087] The secondary load parameter 374 calculated by the strength calculation module 372, together with the corresponding constraints obtained from the constraint module 373, can be used by the optimization problem builder 380 to create a cost function J. The optimization problem builder 380 is understood as an algorithm that creates the cost function J considering the penalty terms of different constraints, their relative relevance (i.e., the weight of each constraint in the cost function), and the corresponding values.

[0088] Furthermore, as also shown in FIG. 5, an optimal trajectory 391 can be found using the optimization problem solver 390. Using an optimization algorithm, an optimal solution, i.e., an optimal set of independent or controllable variables that minimizes the cost function J at the corresponding control steps, can be found. Next, the resulting optimal trajectory 391 from the optimization can be supplied to block 400. In block 400, the first values for each wind turbine actuator can be taken from the calculated optimal trajectory 391 to generate a command 367 for controlling one or more wind turbine actuators of the wind turbine actuator system 364. As shown, FIG. 5 is merely an example of an architecture for implementing the method according to the present disclosure. The details of the controller 360 may be different in other variations. For example, as a non-limiting example, some of the operations defined for different modules can be integrated into a single module.

[0089] In a further embodiment of the present disclosure, the predicted value of the secondary load parameter 374 and / or the determined primary load 375 can be stored and compared with the actually measured values to evaluate the quality of the prediction. Specifically, the quality of the prediction can be characterized, for example, by the average prediction error E m between the predicted value and the actually measured value. The variation E v of the prediction error can also be considered. Both E m and E v can be evaluated over a certain time interval.

[0090] With this additional function, the MPC controller can be appropriately adapted online to the time-varying wind uncertainty and model mismatches, and thus the trade-off between load handling and energy production can be improved. Figure 6 shows an overview of an example of another controller 360’ that can have this function.

[0091] To consider the quality of the prediction, the modified MPC module 362’ can incorporate a strength calculation module monitoring block 410, which can be used to determine the quality of the prediction, for example, for one or more of the secondary load parameter 374 or the primary load 375 to determine E m and E v can be used. For this purpose, the variation of the prediction error E at a certain time interval can be calculated by subtracting the measured value from the predicted value. Then, the obtained value of E is processed to derive E m and E v In particular, the average prediction error E m can be obtained, for example, by averaging E over this time interval or by applying a low-pass filter to the obtained values of E. The variation of the prediction error E v can take various forms (such as the maximum absolute error of E or the standard deviation of E).

[0092] Based on the results of the quality assessment performed by block 410, a strength adaptation module 420 can be provided to adjust different parameters of the optimization problem. The modified parameters 430 can be provided to the optimization problem solver 390.

[0093] In some examples, the quality of the prediction can be used to adjust or modify the value of the secondary load parameter 374.

[0094] Furthermore, as shown above, the cost function can include a slack term for the constraints associated with the quadratic loading parameter 374. The weight of each constraint can indicate the relative importance of the respective quadratic loading parameter 374. In an embodiment of the present disclosure, the weight of the slack component associated with the corresponding quadratic loading parameter 374 can be updated using an evaluation of the quality of the prediction. Thus, in this embodiment, the modified parameter 430 obtained from the adaptation module 420 can correspond to the modified weight of the slack component of the cost function. According to this embodiment, the relative importance of the quadratic loading parameter 374 can be adjusted based not only on the influence of a specific load in a specific wind turbine component, but also on the reliability of the controller itself for appropriately considering the quadratic loading parameter 374.

[0095] In other examples, the quality of the prediction can be used to adjust the constraints, i.e., to adjust the boundary values of the constraints defined for the quadratic loading parameter 374. Thus, the boundaries can be adjusted based on the average prediction error E m and / or the variance E v of the prediction error. Furthermore, the boundary of each constraint can also be adjusted according to the distance between the actually measured value and the currently defined boundary value. As a result, in such an example, the modified parameter 430 can include a modified value of the boundary (either the upper bound UB or the lower bound LB) of the constraints defined for the quadratic loading parameter 374.

[0096] In a specific example, the constrained quadratic loading parameter LP i may be constrained by the original upper bound UB i .

Number

[0097] The constraint can be adjusted with the modified upper bound UB i * .

Number

[0098] Similarly, the original constraint can be included in the cost function J with a certain weight W LPi and this weight W LPi may be adjusted to the modified weight W * LPi .

[0099] The equations for the modified upper limit and the modified weight are given as follows. [Number]

[0100] Here, f(E v , UB i - LP i ) is a non - negative function that adjusts the upper limit of the constraint based on the variance E v of the prediction error, the measured value LP i and the distance between the original upper limit value UB i of the constraint. The function f is a non - decreasing function with respect to E v and a non - increasing function with respect to UB i - LP i . On the other hand, the function g can be a non - negative function that modifies the weight of the original slack penalty based on the variance E v of the prediction error.

[0101] When the constrained quadratic loading parameter 374 is subject to a lower - bound constraint, i.e., when conditions such as LP i >LB i hold (LB i corresponds to the lower - bound value), equivalent processing can be applied except that the equation for the modified boundary is expressed as follows. [Number]

[0102] In other embodiments of the present disclosure, for the primary load 375, processing equivalent to that described for the secondary load parameter 374 can be performed. Therefore, in an example including constraints for the primary load as well, the predicted value and the actual value of the primary load 375 can be used to evaluate the quality of the prediction and adjust the corresponding constraints.

[0103] In another embodiment of the present disclosure, the constraint boundaries defined for different secondary load parameters 374 can be selectively adjusted depending on at least one of the operating state of the wind turbine 10, the operating mode of the wind turbine 10, and environmental conditions. In particular, certain constraints may be more relevant during startup, shutdown, or certain wind conditions (such as gusts, extreme turbulence, or large changes in wind direction). This embodiment of the present disclosure can be suitably adapted to handle such situations. In a variation of this embodiment, a management module configured to provide information regarding, for example, the operating mode of the wind turbine 10 can be included to adjust different constraints.

[0104] In yet another example, the constraints defined for different secondary load parameters 374 can be selectively enabled or disabled according to at least one of the operating state of the wind turbine 10, the operating mode of the wind turbine 10, and environmental conditions. Thus, in this example, not only can the constraints be adjusted according to the conditions, but the constraints can also be removed. Similar to the previous embodiment, in a variation of this embodiment, the management module can be configured to adaptively add or remove constraints based on identifying the current operating state of the wind turbine 10 or the operating mode of the wind turbine 10. This is particularly advantageous in operating modes where certain constraints do not play an important role (for example, constraints related to output during the idling operating mode). By removing unnecessary constraints from the optimization problem, the controller can be made to operate more efficiently.

[0105] Some specific examples regarding different secondary load parameters 374 and advantages of the present disclosure are shown below with reference to FIGS. 7 to 12.

[0106] One of the most important components in the wind turbine 10 is the blade 22. In particular, the blade root portion 24 is known to receive large loads during the operation of the wind turbine. Therefore, a typical controller is known to handle blade loads by imposing constraints on the blade flap moment M flap However, the inventors have found that using the blade flap moment may not represent the true limiting conditions of the blade 22, because the maximum stress may be more accurately characterized by different secondary load parameters.

[0107] Therefore, FIG. 7 schematically shows the combined moment M resultant 510 acting on the blade root portion 24 of the wind turbine blade 22 in the absence of other forces and moments. A cross-sectional view of the blade at the mid-span 221 is also shown (dashed line). Therefore, when the bending moments M x and M y about two orthogonal axes in the blade root portion 24 are given, the combined moment is given by the following equation.

Equation

[0108] FIG. 8 shows the edgewise moment M edge 520, the flap moment M flap 530, and a typical scenario of the combined moment M resultant 510 in the blade root portion 24 of the wind turbine blade 22. The maximum stress calculated when M flap 530 is maximum is smaller than the maximum stress calculated when M resultant 510 is maximum. As a result, M flapIn the constraints based on 530, the worst-case scenario does not always stay within the limit values, and for this reason, blade 22 may be at risk of damage. Therefore, M flap The controller based on 530 may actually indicate that the load is within the limit values even when the load is not, because the true constraint (M resultant 510) is not taken into account.

[0109] Therefore, in an embodiment of the present disclosure, the method can be such that one of the plurality of secondary load parameters includes a combined moment at least at the blade root portion 24 of the wind turbine blade 22. Using the strength calculation module 372, the value of the combined moment M resultant at the blade root 24 is calculated and this combined moment is used as one of the secondary load parameters in the optimization problem, whereby better performance can be obtained than when using M flap . The new constraint can more accurately reflect the true limiting conditions of the wind turbine 10. Next, a constraint can be defined for the combined moment M resultant .

Number

[0110] Here, UB is the upper limit value of the combined moment at the blade root portion 24, and ε is the slack for relaxing the constraint. Therefore, the cost function J of the optimization algorithm can be adapted by adding the corresponding terms. Therefore, in an example of the present disclosure, the cost function can be defined as follows.

Number

[0111] Here, the first four terms are the same as those described above, but new terms are included to account for the soft constraints added to the combined moment in the blade root portion 24. In this case, since a three-blade wind turbine 10 is assumed, MRBi represents the combined moment in blade "i". Further, a weight W is provided to indicate the relative importance of this newly added soft constraint in the optimization problem. εMRB is provided.

[0112] However, since the blade loads can be more complex, in further embodiments of the present disclosure, secondary load parameters can be defined in a more accurate way to capture the true limiting loads. In fact, as shown in FIG. 9, the combined moment M resultant 510 not only affects the stress in the blade root portion 24, but also the axial force F x 513, the shear force S hear 514, and the torsional moment M x 515 in the blade root portion can also be taken into account. As a result, in some embodiments of the present disclosure, one or more of the secondary load parameters among the plurality of secondary load parameters can be derived as a function of a set of three-dimensional forces (F x , F y , F z ) and three-dimensional moments (M x , M y , M z ) at one or more positions of the wind turbine blade 22. More specifically, one or more of the secondary load parameters can include at least one combination of the combination of the combined blade moment 510 in the blade root portion 24 and the axial force 513 in the blade root portion 24, and combinations of normal force, shear force, torsional moment, edge moment, and flap moment at a specific blade portion.

[0113] The combined blade moment M resultant 510 and the axial force F xUsing the combination with 513, the maximum stress in the blade root portion 24 of the wind turbine blade 22 can be explained. Similar to the previous examples, the secondary load parameters can be derived by combining these two different inputs. The corresponding constraints can be formulated as follows.

Number

[0114] Here, both the upper limit value UB and the lower limit value LB are defined. Furthermore, to relax the constraints, slack (ε) is given to both boundaries. Finally, the combined moment M resultant 510 and the axial force F x 513 are given different weights (K1, K2) to reflect the relative importance of M resultant 510 and F x 513 on the stress in the blade root portion 24.

[0115] Other positions of the wind turbine blade 22 other than the root portion 24 may be subject to limit loads depending on the design of the wind turbine 10 and the blade 22. For example, the limit load of the blade 22 in another embodiment of the present disclosure may be the maximum stress of the spar cap in a specific blade portion. In such a case, in order to consider the maximum stress in the spar cap of the wind turbine blade 22, a combination of vertical force, shear force, torsional moment, and flap moment can be determined in such a specific blade portion. More generally, the embodiments of the present disclosure can include determining the resultant force (i.e., a set of three-dimensional forces and three-dimensional moments) in multiple portions of the blade 22. FIG. 10 is a schematic view of the blade 22 showing two portions. In each of these two portions, the following resultant forces R (R1 and R2) can be obtained.

Number

[0116] Here, i = 1...N represents the considered part. The secondary load parameter can be derived as a function of the corresponding resultant force for blade 22.

[0117] Constraints can be included in the relevant secondary load parameters. As an example, the constraints can be included as hard constraints as follows.

Number

[0118] In other examples, the constraints can be relaxed by adding slack variables.

Number

[0119] The examples shown previously related to the secondary load parameters affecting the wind turbine blade 22. However, as can be understood by those skilled in the art and as already explained with reference to FIGS. 1 and 2, the wind turbine 10 includes a very large number of mechanical and structural components. All such components have mechanical limit values, and as a result, several examples of ways to consider secondary load parameters that affect other components can be considered.

[0120] FIG. 11 schematically shows an example of a wind turbine tower 15 including a plurality of parts 151 to 15n, and forces and moments are applied to each part. In the examples of the present disclosure, the secondary load parameter can include a combination of compressive load, bending load, and torsional load in one or more parts of the wind turbine tower 15.

[0121] The combined loads of compression, bending, and torsion on the tower section 151-15n of the wind turbine can also be a limiting factor in the design and operation of the wind turbine 10. Each section of the tower 15 may have different diameters, thicknesses, reinforcements, bolt joints, or materials. Furthermore, the load conditions of each section 151-15n may also be different. Therefore, the limiting loads of a plurality of different tower sections may be different. The limiting load of each section can be a combination of the maximum stress, maximum deformation, or buckling load. Thus, in one embodiment of the present disclosure, the constraints of the tower can be defined based on a plurality of secondary load parameters defined for the limiting load of each section. The secondary load parameters can be calculated by the strength calculation module 372 so that an accurate representation of the strength of the tower 15 during operation can be obtained.

[0122] As another example of the present disclosure, the secondary load parameters can include at least one of the maximum tensile stress, maximum shear stress, and maximum shear deformation energy at one or more positions of the wind turbine hub 20. More specifically, the maximum tensile stress, maximum shear stress, and maximum shear deformation energy can be calculated using a stress tensor after considering the three-dimensional forces and moments from each wind turbine blade 22 connected to the hub 20 and the three-dimensional forces and moments from the connection to the main shaft 44 of the wind turbine 10.

[0123] Regarding this, FIG. 12 shows a schematic diagram of the wind turbine hub 20 and the corresponding loads. External forces and moments are applied to the hub 20 by each of the three pitch bearings 72 of the pitch bearing and by the connection to the main shaft 44. A resultant force including three-dimensional forces and moments is obtained for each of these positions. Next, the strength calculation module 372 can calculate the corresponding secondary load parameters at specific positions of the hub 20. The positions may correspond to the positions where the largest loads are received in the hub 20. These positions can be determined in advance, for example, by analysis using FEM. For each predetermined position in the hub 20, the secondary load parameters are derived, and the corresponding constraints can be defined as follows. [Number]

[0124] Here, R bladei is the resultant of the three-dimensional forces and moments at the corresponding pitch bearing connection to the hub 20, and R MainShaft is the resultant of the three-dimensional forces and moments with respect to the connection to the main shaft 44. The said forces and moments can be obtained using a simple aerodynamic elastic model.

[0125] In this case, the strength calculation module 372 can calculate the limiting conditions of the hub material at the selected positions (a, b, c in FIG. 12). For the purpose of performing this calculation, the resultant of the three blades 22 and the main shaft 44 can be supplied to the strength calculation module 372. The strength calculation module 372 can include the local stress tensor at the selected predetermined position. The limiting conditions can be derived from the resultant force R and the stress tensor. In the examples of the present disclosure, different limiting conditions can be selected considering the different properties of the materials. Therefore, the maximum tensile stress, the maximum shear stress, or the maximum shear deformation energy can be selected. In any case, the maximum stress σ max can be defined as the upper limit of the selected stress. In an example including a ductile material, σ max can correspond to the yield strength σ yield . This is the case of the example shown in FIG. 12, and this maximum tension corresponds to the yield stress limit σ yield of the hub material, and the corresponding constraints can be defined using this. As will be understood by those skilled in the art, different limiting conditions can be defined for different components based on the corresponding materials.

[0126] According to another aspect of the present disclosure, a controller 360 for the wind turbine 10 is provided. The controller 360 is configured to execute a method for controlling the wind turbine 10 according to the method example described above or other equivalent examples. The controller 360 can be implemented as part of the wind turbine controller 36.

[0127] According to another aspect of the present disclosure, a wind turbine 10 may be provided that includes a tower 15, a nacelle 16 rotatably attached to an upper portion of the tower 15, a wind turbine rotor 18 including a plurality of wind turbine blades 22, and a yaw system. The wind turbine can include a control system 360 configured to perform in a manner as described throughout the present disclosure or any equivalent example (including combinations).

[0128] The description provided herein discloses the teachings using examples, and any person skilled in the art can implement the teachings (including manufacturing and using the device or system and performing the incorporated methods). The patentable scope is defined by the claims and can 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 have structural elements that do not differ from the literal language of the claims or include equivalent structural elements that do not have a substantial difference from the literal language of the claims. Aspects of the various described embodiments, as well as other known equivalents of such aspects, can be mixed and combined by those skilled in the art to construct additional embodiments and techniques that follow the principles of this application. If reference signs related to the drawings are shown in parentheses in the claims, the signs are merely for facilitating the understanding of the claims and should not be construed as limiting the scope of the claims.

Description of Reference Signs

[0129] 10 Wind turbine

Claims

1. A method (100) for controlling a wind turbine (10) having a plurality of wind turbine actuators (364), comprising: receiving operating data (366) of the wind turbine (10); estimating an operating state of the wind turbine (10) based on the received operating data (366); predicting a possible operating state of the wind turbine (10) according to the operation of the wind turbine actuator (364) over a finite time using a control model (370), wherein the control model (370) includes a predictive aeroelastic control model (371) for determining a primary load (375) based on the received operating data (366), and the control model (370) further includes a strength calculation module (372) for calculating one or more secondary load parameters (374) based on one or more primary loads (375), and constraints are defined for the secondary load parameters (374); optimizing a cost function over a finite optimization time according to the constraints to determine an optimal trajectory (391) including commands for the wind turbine actuator (364); controlling the wind turbine actuator (364) using the first command (367) of the determined optimal trajectory A method comprising.

2. The method according to claim 1, further comprising repeating the steps at consecutive control intervals.

3. Constraints are defined for one or more of the plurality of primary loads (375), and optimizing the cost function over a finite optimization time also complies with the constraints defined for the primary loads (375). The method according to claim 1 or 2.

4. The secondary load parameter (374) is included in the cost function. Specifically, the secondary load parameter (374) is included with a corresponding weight indicating the relative importance of the secondary load parameter (374). More specifically, the constraints defined for the secondary load parameter (374) include at least an upper limit value and a slack variable, and the slack variable is penalized in the cost function. The method according to any one of claims 1 to 3.

5. The method according to any one of claims 1 to 4, wherein the constraint defined for the secondary load parameter (374) is based on a limit value of a material of a component of one or more wind turbines (10).

6. The secondary load parameter (374) is predetermined. Specifically, the secondary load parameter (374) is predetermined based on a finite element method analysis of a selected component of the wind turbine (10). The method according to any one of claims 1 to 5.

7. The secondary load parameter (374) includes at least one of force, moment, stress, strain, or buckling load. More specifically, the secondary load parameter (374) includes stress in a component of the wind turbine derived from a stress tensor, or strain of a component of the wind turbine derived from a strain tensor. The method according to any one of claims 1 to 6.

8. The predicted value and / or the determined primary load of the secondary load parameter (374) are stored, compared with the actually measured value, and the quality of the prediction is evaluated. More specifically, the quality of the prediction is characterized by an average prediction error over a certain time interval and / or by a variation of the prediction error over a certain time interval. The method according to any one of claims 1 to 7.

9. The secondary load parameter (374) is included in a cost function together with a corresponding weight indicating the relative importance of the secondary load parameter (374). Using the evaluation of the quality of the prediction, the weight of the corresponding secondary load parameter (374) is updated and / or the boundary of the constraint defined for the corresponding secondary load parameter (374) is updated. The method according to claim 8.

10. The constraints defined for different secondary load parameters (374) are selectively adjusted depending on at least one of the operating state of the wind turbine (10), the operating mode of the wind turbine (10), and environmental conditions. More specifically, the constraints defined for different secondary load parameters (374) are selectively enabled or disabled according to at least one of the operating state of the wind turbine (10), the operating mode of the wind turbine (10), and environmental conditions. The method according to any one of claims 1 to 9.

11. One of the plurality of secondary load parameters (374) includes at least the combined moment at the blade root portion (24) of the wind turbine blade (22), the method according to any one of claims 1 to 10.

12. One or more secondary load parameters are derived as a function of a set of three - dimensional forces (F ｘ , F ｙ , F ｚ ) and three - dimensional moments (M ｘ , M ｙ , M ｚ ) at one or more positions of a wind turbine blade (22), and more specifically, one or more secondary load parameters (374) are A combination of the combined blade moment (510) and the axial force (513), and A combination of the normal force, shear force (514), torsional moment (515), edge moment, and flap moment at a specific blade (22) portion The method according to any one of claims 1 to 11, including at least one of.

13. The secondary load parameter (374) includes a combination of compressive load, bending load, and torsional load in one or more parts of the wind turbine tower (15), the method according to any one of claims 1 to 12.

14. The secondary load parameter (374) includes at least one of the maximum tensile stress, maximum shear stress, and maximum shear deformation energy at one or more positions of the wind turbine hub (20). More specifically, the maximum tensile stress, maximum shear stress, and maximum shear deformation energy are calculated using the stress tensor after considering the three-dimensional forces and moments from each wind turbine blade (22) connected to the hub (20) and the three-dimensional forces and moments from the connection to the main shaft (44) of the wind turbine (10), the method according to any one of claims 1 to 13.

15. A controller (360) of the wind turbine (10), wherein the control system (360) is configured to execute the method according to any one of claims 1 to 14, the controller (360).

Images