• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    "Method and System for Optimizing Wind Farm Operation and Method for Operating a Wind Farm" These are presented achievements of the methods and systems for optimizing the operation of a wind farm. The method includes receiving new values corresponding to at least some swirl parameters for wind turbines in the wind farm. The method additionally includes identifying new series of interacting wind turbines from wind turbines based on the new values. In addition, the method includes developing a park-level prediction swirl model for the new series of interacting wind turbines based on the new values and historical swirl models determined using historical values of the corresponding series swirl parameters. wind turbine frameworks interacting in the wind farm. Additionally, the method includes adjusting one or more control settings for at least the new series of interacting wind turbines based on the park-level forecast swirl model.
    公开号:BR102015009550A2
    申请号:R102015009550-3
    申请日:2015-04-28
    公开日:2018-03-06
    发明作者:Krishnamurty Ambekar Akshay；Menon Anup；Vikrambhai Desai Kalpit；Dattaram Dhuri Krishnarao；Chandrashekar Siddhanth
    申请人:General Electric Company；
    IPC主号:
    专利说明:

    (54) Title: METHOD AND SYSTEM FOR OPTIMIZING THE OPERATION OF A WIND FARM AND METHOD FOR OPERATING A WIND FARM (51) Int. Cl .: F03D 7/04; F03D 9/25; F03D 17/00; F03D 9/00; G05B 13/04 (52) CPC: F03D 7/045, F03D 7/048, F03D 9/257, F03D 17/00, F03D 9/00, G05B 13/048 (30) Unionist Priority: 4/29/2014 IN 2155 / CHE / 2014 (73) Holder (s): GENERAL ELECTRIC COMPANY (72) Inventor (s): AKSHAY KRISHNAMURTY AMBEKAR; ANUP MENON; KALPIT VIKRAMBHAI DESAI; KRISHNARAO DATTARAM DHURI; SIDDHANTH CHANDRASHEKAR (74) Attorney (s): CAROLINA NAKATA (57) Summary: METHOD AND SYSTEM FOR OPTIMIZING THE OPERATION OF A WIND FARM AND METHOD FOR OPERATING A WIND FARM These are the presented achievements of the methods and systems to optimize the operation of a wind farm. The method includes receiving new values corresponding to at least some eddy parameters for wind turbines in the wind farm. The method additionally includes identifying new series of interacting wind turbines from the wind turbines based on the new values. In addition, the method includes developing a forecast swirl model at the park level for the new series of wind turbines in interaction based on the new values and historical swirl models determined using the historical values of the swirl parameters corresponding to the series references of wind turbines interacting in the wind farm. In addition, the method includes adjusting one or more control settings for at least the new series of wind turbines in interaction based on the forecast swirl model at the park level.
    1/41 «z“ METHOD AND SYSTEM FOR OPTIMIZING THE OPERATION OF A WIND FARM AND METHOD FOR OPERATING A WIND FARM ”
    Background [001] The achievements of the present disclosure relate, in general, to wind turbines and, more particularly, to methods and systems to optimize the operation of a wind farm.
    [002] Renewable energy resources are increasingly used as the purest and most cost-effective alternatives to fossil fuels to supply global energy requirements. Wind energy, in particular, has emerged as one of the most favorite renewable energy resources as it is abundant, renewable, widely distributed and pure. In general, wind energy can be harnessed by wind turbines that are designed to produce electrical energy in response to a wide spectrum of wind speeds. These wind turbines are typically located in a wind farm spread over a specific geographic region so that the wind that passes through that region causes the blades associated with the wind turbines to rotate. Each of the rotating blades, in turn, causes a rotor of an associated generator to turn, which helps in the generation of electrical power.
    [003] Traditionally, wind farms are controlled in a decentralized manner to generate power so that each turbine is operated to maximize local power output and to minimize the impacts of local fatigue and extreme loads. However, in practice, such independent optimization of wind turbines ignores performance targets at the park level, thereby leading to under-ideal performance at the level of the wind farm. For example, independent optimization of wind turbines may not be responsible for aerodynamic interactions, such as swirl effects between turbines close together within the wind farm that can
    2/41 affect a power output at the park level.
    [004] Typically, whirlpool effects include a reduction in wind speed and increased wind turbulence in a downstream wind turbine due to conventional upstream wind turbine operation. The reduced wind speed causes a proportional reduction in a power output from the downstream wind turbine. In addition, the increased turbulence increases the fatigue loads placed on the downstream wind turbine. Several studies have reported a loss of more than 10% in the annual energy production (AEP) of the wind farm due to swirl effects between nearby wind turbines optimized independently within the wind farm.
    [005] Consequently, some approaches currently available attempt to optimize power generation at the level of the wind farm by mitigating an impact of whirlpool effects through coordinated control of wind turbines in the wind farm. Typically, mitigating the eddy effects involves precisely modeling the eddy effects suffered by different wind turbines in the wind farm. For example, empirical or semi-empirical models based on buoyancy and / or based on high fidelity physics can be used to model the whirlpool effects between wind turbines interacting aerodynamically in the wind farm.
    [006] Conventionally, empirical or semi-empirical models (engineering whirlpool models) are generated based on field experiment data and / or historical wind information. Consequently, these models can be used to design the wind farm layouts in order to optimize one or more performance targets before installing the wind turbines. Alternatively, these models can be used to optimize the performance of the wind farm
    3/41 subsequent to installation.
    [007] An optimization approach, for example, employs engineering whirlpool models to determine control definitions for wind turbines. In particular, the engineering whirlpool models determine the control settings in order to operate the upstream turbines at lower efficiencies which, in turn, allow for better energy recovery in the downstream turbines. Another approach uses engineering swirl models to adjust an upstream turbine yaw alignment in relation to an approaching wind direction in order to remove the resulting swirl effects from the downstream turbines.
    [008] However, conventional engineering models are not responsible for the predominant wind influx and other environmental conditions, such as atmospheric boundary layer stability and adequate longitudinal turbulence intensity. Since environmental conditions in the wind farm tend to change frequently, the eddy models estimated using engineering eddy models may be inaccurate for use during a real-time deployment. Inaccurate modeling of whirlwind conditions, in turn, can result in the use of incorrect control settings for wind turbines in the wind farm. In this way, conventional optimization approaches using engineering swirl models usually provide only marginal improvement in performance output at the park level.
    [009] Consequently, high-fidelity eddy models, for example, based on computational fluid dynamics modeling have been explored to provide better accuracy in modeling eddy interactions. High-fidelity models result in measurements and analyzes of a wide variety of parameters that
    4/41 require additional instrumentation, complex computations and associated costs. The cost and complexity associated with high-fidelity models, therefore, may prohibit the widespread use of these models in every turbine in the wind farm and / or for real-time optimization of wind farm operations.
    Brief Description [010] In accordance with one aspect of the present disclosure, a method for optimizing the operation of a wind farm is presented. The method includes receiving new values corresponding to at least some eddy parameters for wind turbines in the wind farm. The method additionally includes identifying new series of interacting wind turbines from the wind turbines based on the new values. In addition, the method includes developing a forecast swirl model at the park level for the new series of wind turbines in interaction based on the new values and historical swirl models determined using the historical values of the swirl parameters corresponding to the series references of wind turbines interacting in the wind farm. In addition, the method includes adjusting one or more control settings for at least the new series of wind turbines in interaction based on the forecast swirl model at the park level.
    [011] In accordance with an additional aspect of the present disclosure, another method for operating a wind farm is disclosed. The method includes assembling the historic swirl models for different series of interacting wind turbines in the wind farm based on historical values from selected combinations of swirl parameters corresponding to the series of interacting wind turbines. In addition, the method additionally includes determining the ideal control settings for each wind turbine in the series of interacting wind turbines for each of the wind turbines.
    5/41 selected combinations of swirl parameters based on historical swirl models. In addition, the method includes storing the ideal control settings for each wind turbine as a function of the selected combination of eddy parameters. The method additionally includes receiving the new values of the eddy parameters obtained in a subsequent period of time after obtaining the historical values. In addition, the method includes determining the control settings for wind turbines in each of the new series of wind turbines using new values and stored control settings.
    [012] In accordance with yet another aspect of the present disclosure, a system to optimize the operation of a wind farm is presented. The system includes a plurality of wind turbines, one or more monitoring devices configured to measure the values of a plurality of eddy parameters for one or more of the plurality of wind turbines and a park control subsystem operationally coupled to at least least one of the monitoring devices. The park control subsystem is programmed to receive new values corresponding to at least some eddy parameters for the plurality of wind turbines in the wind farm. The park control subsystem is additionally programmed to identify the new series of wind turbines in interaction from the plurality of wind turbines based on the new values. In addition, the park control subsystem is programmed to develop a forecast swirl model at the park level for the new series of wind turbines in interaction based on the new values and historical swirl models determined using the historical values of the wind turbines. eddy parameters corresponding to the reference series of wind turbines interacting in the wind farm. Additionally, the park control subsystem is
    6/41 programmed to adjust one or more control settings for at least the new series of wind turbines in interaction based on the forecast swirl model at the park level.
    Drawings [013] These and other features and aspects of realizations in the present disclosure will be better understood when the following detailed description is read with reference to the accompanying drawings in which similar characters represent similar parts throughout the drawings, where:
    - Figure 1 is a diagrammatic illustration of an exemplary wind farm, according to the achievements of the present disclosure;
    - Figure 2 is a flowchart that illustrates an exemplary method for optimizing the operation of a wind farm, in accordance with the achievements of the present disclosure;
    - Figure 3 is a schematic representation of an exemplary sequence to determine the ideal control definitions for the series of wind turbines interacting aerodynamically in a wind farm, in accordance with the achievements of the present disclosure;
    - Figure 4 is a flowchart that illustrates an exemplary method for optimizing the operation of a wind farm in a delayed optimization mode, according to the achievements of the present disclosure;
    - Figure 5 is a flowchart that illustrates an exemplary method for optimizing the operation of a wind farm in real time, according to the achievements of the present disclosure; and
    - Figure 6 is a graphical representation that depicts a comparison of energy gains achieved with the use of different whirlpool models, according to the achievements of the present disclosure.
    Detailed Description [014] The following description presents exemplary achievements
    7/41 of systems and methods to optimize the operation of a wind farm. In particular, the achievements illustrated in this document reveal a method for modeling driven by swirl effects data using data conventionally aggregated by a Supervisory Control and Data Acquisition (SCADA) server at the wind farm. Aggregated data may include whirlpool parameters that include environmental conditions, wind farm geometric layout and / or operational information corresponding to wind turbines. At least some of the swirl parameters, such as the geometric layout, can be recognized or received once, while some other swirl parameters, such as environmental conditions, can be monitored continuously to help estimate the effects of whirlpool at the level of the wind farm.
    [015] Environmental conditions may, for example, include predominant wind direction, wind speed detected in an upstream wind turbine (upstream wind speed), wind speed detected in a downstream wind turbine (downstream wind speed) downstream), sharp wind, wind rotation, temperature, humidity, and / or pressure. Eddy parameters can additionally include operational information and control settings, such as a peak speed ratio, a pitch angle, a yaw alignment, a generator speed, a power output, a torque output, a thrust measurement, and / or operating states of individual wind turbines that provide information regarding any wind turbines in the wind farm that are not producing power. In addition, the eddy parameters can also include a known geometric layout of the wind farm that includes information corresponding to the wind farm's terrain, the number of turbines close together, the actual turbine locations and / or
    8/41 relative locations of downstream and upstream wind turbines.
    [016] In addition, the achievements of the present disclosure present a data driven approach that uses the monitored values of eddy parameters to generate robust eddy models at park level. Specifically, the data-driven approach uses eddy parameters to aerodynamically identify the series of interacting wind turbines and estimate the corresponding eddy interactions (in series). Serial swirl interactions, in turn, are used to generate swirl models at the park level in real time. The use of prevailing environmental conditions and the current operating states of individual turbines enables the detection of whirlpool interactions that are suffered in real time by the downstream wind turbines, thereby allowing the determination of a forecast swirl model at the park level. more accurate.
    [017] Additionally, the determination of the eddy model at the park level through a series assessment of eddy interactions reduces computational effort, thus allowing the real-time optimization of one or more performance targets selected for the park wind power. Specifically, the whirlpool model at the park level assists in determining ideal control settings for the different wind turbines in the wind farm in order to significantly optimize overall performance targets, such as maximizing annual energy production (AEP) and minimize the fatigue loads suffered by wind turbines in the wind farm.
    [018] Although the exemplary achievements of the present systems and methods are described in the context of optimizing different performance targets for a wind farm, it will be noted that the use of the achievements of the present system in various other applications is also
    9/41 contemplated. By way of example, certain embodiments of the present disclosure can be employed to optimize the operations of a plurality of tidal or hydroturbine turbines or in submerged systems. An exemplary environment that is suitable for practicing various deployments of the present system is discussed in the following sections with reference to Figure 1.
    [019] Figure 1 illustrates an exemplary wind farm 100 according to aspects of the present disclosure. In one embodiment, wind farm 100 includes a plurality of wind turbines 102 arranged in a desired geometric layout. For example, wind turbines 102 can be arranged randomly, in a single row, or in an array of rows and columns using one or more layout optimization algorithms. In general, optimization algorithms can be designed to maximize the positive effects of expected wind speed and direction on performance targets, for example, AEP, while minimizing the negative effects of such an increase in associated fatigue loads to each of the individual wind turbines 102.
    [020] In one embodiment, each of the wind turbines 102 includes one or more energy conversion modules, such as rotor blades 104, a lifting gearbox (not shown) and a power generator (not shown) which converts wind energy into usable electrical energy. In addition, wind turbines 102 also include blade pitch mechanisms (not shown) to regulate the turbine power output and rotor speed, yaw mechanisms (not shown) and one or more monitoring devices 110 that work in a manner cohesive with other wind turbine equipment 102 to rotate and align rotor blades 104 in line and / or in relation to the prevailing wind direction. In addition, wind turbines 102 can also include
    10/41 cooling (not shown) to prevent wind turbine components 102 from overheating, braking systems (not shown) to stop rotor blades 104 from turning when desired, and nacelles (not shown) to protect the different components of wind turbines 102 from environmental factors.
    [021] Typically, the rotor blades 104 of wind turbines 102 are aligned in a substantially similar direction, for example, the approaching wind direction during the operation of wind turbine 102. Such blade alignment, however, positions certain downstream wind turbines 102 behind certain upstream wind turbines 102 in wind farm 100, thus resulting in whirlwind effects that otherwise impact the operations of downstream wind turbines 102. For example, the wind that blows in rotor blades 104 of upstream wind turbines 102 causes the corresponding blades 104 to rotate. The rotating blades 104 convert at least part of the oncoming kinetic energy into mechanical energy, thus reducing the wind speed suffered by the wind turbine downstream 102, while additionally increasing turbulence.
    [022] Since the power output of wind turbines 102 is proportional to the approaching wind speed, a reduction in the wind speed in the downstream wind turbine 102 due to eddy effects reduces a corresponding power output. Additionally, turbulence caused by swirling effects can damage turbine components due to cyclic fatigue loading. For example, the fatigue load can initiate small cracks in surfaces of the turbine components that can increase in size and thus propagate, potentially leading to the failure of the downstream wind turbine 102.
    [023] Furthermore, since the independent optimization of
    11/41 wind turbines can further aggravate the whirlpool effects, it is desirable to configure the operation of the wind turbines 102 so that the power output at the park level, the AEP, and / or the fatigue loads in the wind farm 100 remain within corresponding designated thresholds. In particular, it is desirable to continue to adjust the control settings for each of the interacting wind turbines 102 based on swirl parameter variation values, such as real-time wind speed and direction so that performance targets are met. park level are achieved consistently.
    [024] Consequently, each of the wind turbines 102 includes one or more turbine controllers 106 that regulate operation of the corresponding wind turbines 102 to mitigate the swirling effects between the interacting wind turbine series 102. In one embodiment, the wind turbine controllers turbine 106 regulate the operation of wind turbines 102 based on environmental conditions, user inputs and / or commands received from an associated park control subsystem 108. Consequently, turbine controllers 106 may include application-specific processors, a logic controller programmable (PLC), digital signal processors (DSPs), microcomputers, microcontrollers, Application Specific Integrated Circuits (ASICs) and / or Field Programmable Port Arrays (FPGAs).
    [025] Additionally, turbine controllers 106 can be communicatively coupled to park control subsystem 108 and / or to a plurality of monitoring devices 110 via a wireless and / or wired communications network 112. A Communications network 112, for example, may include the Internet, a local area network (LAN), wireless local area networks (WLAN), wide area networks (WAN), such as, Global Inoperability networks for Access by Microwave (WiMax), networks
    12/41 satellite, cellular networks, sensor networks, Ad-hoc networks and / or short-range networks.
    [026] In addition, monitoring devices 110, for example, include encoders or sensors that provide direct or indirect measurement of eddy parameters, such as wind speed, wind direction, ambient temperature, pressure, density, turbulence, wind shear, and / or power output from wind turbines 102. In certain embodiments, monitoring devices 110 may be positioned inside and / or outside wind farm 100 to measure eddy parameters, such as SCADA information that include the suffered and / or expected wind in different wind turbines 102. In one embodiment, for example, monitoring devices 110 may be arranged on or near wind turbines 102 to measure SCADA information corresponding to environmental conditions . SCADA information can be used by turbine controllers 106 and / or the park control subsystem 108 to continuously estimate aerodynamic interactions between interacting wind turbine series 102. Estimated aerodynamic interactions or eddy effects, in turn, they can be used to determine the ideal control settings for the definitions of the interacting wind turbines 102 in real time.
    [027] In one embodiment, monitoring devices 110 can be configured to store SCADA information in a storage repository 114 for future processing. For this purpose, the storage repository 114 can be communicatively coupled to the turbine controllers 106, the park control subsystem 108 and / or the monitoring devices 110 via the communications network 112. In addition, the storage repository 114, for example, includes one or more hard drives, floppy drives,
    13/41 rewritable compact disc drives (CD-R / W), Digital Versatile Disc drives (DVD), flash drives, optical drives and / or solid state storage arrangement to store SCADA information.
    [028] Alternatively, monitoring devices 110 can be configured to communicate SCADA information to turbine controllers 106 and / or to park control subsystem 108 at one or more designated time intervals. In certain additional embodiments, monitoring devices 110 can be configured to communicate SCADA information to turbine controllers 106 and / or to park control subsystem 108 at random intervals upon receipt of a user request and / or upon the determination of a dignified change (for example, a change greater than 5%) in consecutive measurements.
    [029] In certain embodiments, park control subsystem 108 can be configured to use SCADA information received from turbine controllers 106 to supervise and / or control the operations of turbine controllers 106 and / or wind turbines 102. For this purpose, the park control subsystem 108 may include, for example, one or more application specific processors, DSPs, microcomputers, microcontrollers, PLCs, ASICs and / or FPGAs. Although Figure 1 illustrates the park control subsystem 108 as a single centralized server, in an alternative, the park control subsystem 108 can correspond to a distributed system.
    [030] In addition, in one embodiment, the park control subsystem 108 uses the SCADA information along with a geometric layout of the wind farm 100 to model whirlpool interactions with intermediate turbine based on prevailing environmental conditions. For example, park control subsystem 108 may employ a
    14/41 whirlwind modeling driven by data that fits a prediction progression model for SCADA information that is conventionally aggregated in wind farm 100. Typically, a regression model defines a statistical relationship that can be used to indicate a change in a dependent variable when one or more independent variables are varied, while other independent variables are kept fixed. However, a predetermined statistical relationship between the eddy parameters may not be valid for new and subsequently obtained values from the eddy parameters.
    [031] Consequently, in one embodiment, the predetermined statistical relationship can undergo machine learning and subsequent validation with additional swirl parameter values. Specifically, in certain achievements, training and evaluation continue until an accurate characterization of the eddy interactions in real time between the series of interacting wind turbines 102 with the use of the statistical relationship reaches a permanent regime. Once the steady state is achieved, the park control subsystem 108 identifies aerodynamically interacting wind turbine series from SCADA information obtained recently and uses the statistical relationship to determine the series eddy models for each series of wind turbines upstream and downstream 102 at wind farm 100.
    [032] In certain embodiments, serial whirlpool models can be configured to modulate an optimization problem for turbine control definitions based on one or more forecast variables, desired performance goals and / or known restrictions. For example, in one embodiment, the regression model can assist in determining optimal control settings for wind turbines 102 based on a wind turbine 102 operating regime. Typically, in
    15/41 low wind speeds, wind turbines 102 operate in a variable speed mode, while operating at a nominal speed and in a power mode at high wind speeds. The park control subsystem 108 can determine the operating regime based on information from the current wind, turbine rotor speed, pitch angle and / or the power collected as part of the SCADA information.
    [033] Based on the operation regime, in one realization, the park control subsystem 108 can be configured to determine ideal values of one or more forecast variables corresponding to the performance targets having one or more known operational restrictions. As used in this document, the term “forecast variables” can be used to refer to values that can be manipulated in order to arrive at an ideal performance target value and at the same time that it satisfies operational restrictions. In one embodiment, the forecast variables include the control settings for wind turbines 102, such as a peak speed ratio and blade pitch angles. In general, different definitions of forecast variables can be used to influence the behavior of wind turbine 102 under different regimes and operating conditions.
    [034] For example, during operation in variable speed mode, the regression model can use the short pitch angle and / or peak speed ratio definition points as the forecast variables. As used in this document, the term “short pitch angle” definition point corresponds to the pitch angle at which rotor blades 104 are locked during variable speed mode and the term “speed ratio definition point” tip ”corresponds to control set points used to achieve a peak speed ratio
    16/41 required during turbine operation. In general, the peak velocity ratio can be defined as the ratio of a linear velocity of the blade tip to a wind speed equivalent to the power.
    [035] However, during operation of wind turbines 102 in rated power mode, park control subsystem 108 can use the turbine power set point and the rotor speed set point as the forecast variables . In one embodiment, the forecast variables for the regression model help to determine the performance of a wind turbine, such as its power output, fatigue loads and downstream eddy effects in view of the prevailing wind conditions.
    [036] In general, the whirlwind effects suffered by a downstream turbine 102 result not only from the operation of a corresponding upstream turbine 102, but also from the operation of other wind turbines 102 and surround the land at wind farm 100. Typically, Due to the size and layout of wind farm 100, the eddy effects cascade from an upstream wind turbine 102 to two or more downstream wind turbines 102 that are located in the oncoming wind path. As a result, park control subsystem 108 can develop a forecast swirl model at the park level based on the swirl models determined in series to provide a more comprehensive estimate of the overall swirl effects in wind farm 100. As previously verified, the use of prevailing environmental conditions and operating states of individual turbines allows the capture of whirlpool interactions that are experienced in real time in downstream wind turbines, thereby allowing the determination of a whirlpool model at the level of the most accurate park trustworthy. Additionally, the determination of the whirlpool model at the park level
    17/41 through a series assessment of eddy interactions reduces computational effort, thus allowing real-time optimization of one or more selected performance targets for the wind farm. Certain exemplary embodiments of methods for determining the series vortex models and the park level forecast vortex model are described in more detail with reference to Figures 2 to 5.
    [037] In one embodiment, the eddy model at the park level can be used to predict expected eddy interactions between the series of interacting wind turbines 102 for the prevailing environmental conditions and different combinations of control settings, such as the pitch angle and / or peak speed ratio set point. Consequently, in one embodiment, the whirlpool model at the park level can be used to determine and adjust one or more control settings for each of the wind turbines interacting aerodynamically 102. The control settings include, for example, a definition of peak speed ratio, a yaw misalignment, a short pitch definition point and / or a rotor speed definition point. Specifically, the park control subsystem 108 adjusts the control settings for one or more wind turbines 102 in the wind park 100 in order to achieve one or more desired performance targets. For example, in one embodiment, the park control subsystem 108 can adjust the pitch angle of a rotor blade, change a generator torque, change a generator speed, change a nacelle yaw, brake one or more components wind turbine, additional or activates an airflow and / or modifying an element on a rotor blade surface to activate the desired performance targets.
    [038] Particularly, in one realization, the subsystem of
    18/41 park control 108 uses the park level forecasting whirlpool model to adjust wind turbine control settings 102 to maximize power output at park level and / or AEP in view of varying conditions environmental issues. In another example, park control subsystem 108 uses the park-level forecast vortex model to selectively adjust wind turbine control settings 102 to minimize turbine fatigue loads. Alternatively, park control subsystem 108 uses the park-level forecasting whirlpool model to selectively adjust one or more wind turbine control settings 102 for restricted optimization of desired performance goals, such as optimizing AEP, while maintaining fatigue loads on individual wind turbines 102 below a designated threshold. Certain exemplary achievements of methods for adjusting the control settings of wind turbines 102 to optimize one or more performance targets based on the whirlpool model at the level of the precision park are described in more detail with reference to Figures 2 to 5.
    [039] Figure 2 illustrates a flowchart 200 that depicts an exemplary method for optimizing the operation of a wind farm. In the present specification, the achievements of the exemplary method can be described in a general context of non-transitory computer executable instructions on a computer system or on a processor. In general, computer-executable instructions can include routines, programs, objects, components, data structures, procedures, modules, functions, and the like that perform particular functions or implant abstract types of particular data.
    [040] Additionally, the achievements of the exemplary method can also be practiced in a computing environment distributed across
    19/41 that the optimization functions are performed by remote processing devices that are connected via a wired and / or wireless communication network. In the distributed computing environment, executable instructions per computer can be located on both local and remote storage media, including memory storage devices.
    [041] Additionally, in Figure 2, the example method is illustrated as a collection of blocks in a logical flowchart, which represents the operations that can be implemented in hardware, software or in combinations thereof. The various operations are depicted in the blocks to illustrate the functions that are performed, for example, during the steps to receive one or more historical swirl models, develop a forecast swirl model at the park level and / or adjust one or more settings control in the example method. In the software context, the blocks represent computer instructions that, when executed by one or more processing subsystems, perform the recited operations.
    [042] The order in which the exemplary method is described is not intended to be interpreted as a limitation, and any number of the described blocks can be combined in any order to implement the exemplary method disclosed in this document or an equivalent alternative method. In addition, certain blocks can be deleted from the example method or augmented by additional blocks with added functionality without departing from the spirit and scope of the matter described in this document. For purposes of discussion, the example method will be described with reference to the elements in Figure 1.
    [043] As previously verified, upstream turbine operation reduces wind speed and increases the intensity of turbulence
    20/41 in wind turbines downstream in a wind farm, such as wind farm 100 in Figure 1. Such aerodynamic interactions between upstream and downstream wind turbines correspond to eddy effects, which typically reduce power output and the service life of the downstream wind turbine components. Conventional techniques employ high fidelity models or simplified engineering eddy models based on physics or empirical data to model eddy interactions between different wind turbines in the wind farm. The estimated eddy interactions are then used for the adapted control settings, such as a rotor speed and / or upstream wind turbine blade alignment. The cost and computational costs associated with high-fidelity models, however, prohibit the use of high-fidelity models on a large scale. In addition, since conventional engineering models ignore information corresponding in real time to prevailing unformed environmental conditions and / or turbine performance, engineering whirlpool models are able to provide only a marginal improvement in the performance goals of a wind farm.
    [044] In contrast, the achievements of the present disclosure present an exemplary method for accurately modeling swirl effects at the park level based on a regression model driven by recursive data to provide superior performance compared to conventional wind turbine operations . The method starts at step 202, where the historical values of at least some whirlpool parameters for wind turbines in a wind farm are received. In one embodiment, the historical values correspond to the monitored values of eddy parameters aggregated by a SCADA server, such as, the park control subsystem 108 of Figure 1 for
    21/41 • I a designated period of time, for example, a few days, a few hours or a few minutes. Whirlpool parameters, as previously verified, include environmental conditions, control settings for individual wind turbines, geometric layout of the wind farm and / or any parameter that may affect the operational efficiency of individual wind turbines and / or the wind farm generally. In certain embodiments, only certain eddy parameters, such as upstream wind speed, downstream wind speed, wind direction, control settings, and / or turbine operating or out of operation states are received continuously. Other eddy parameters, such as the wind farm's geometric layout can be received only once or can be retrieved from an associated storage device, such as the storage repository 114 in Figure 1.
    [045] Additionally, in step 204, a reference series of interacting wind turbines is identified from the wind turbines based on historical values. Particularly, in one embodiment, a park control subsystem, such as park control subsystem 108 in Figure 1, can be configured to identify the reference series of wind turbines in interaction based on the prevailing wind direction and geometric layout of the wind farm. In addition, the park control subsystem uses the geometric layout of the wind farm to determine the nearby turbines and / or the relative positions of the wind turbines for a detected wind direction. The relative positions, in turn, may allow the identification of series of interacting wind turbines so that each series includes at least one upstream wind turbine and at least one downstream wind turbine that suffers whirlwind effects.
    [046] Alternatively, in certain embodiments, the
    22/41 The park control subsystem can employ an engineering whirlpool model, such as the Jensen or Ainslie model to identify aerodynamically interacting wind turbine series. In general, the Jensen or Ainslie models can identify the reference series of interacting wind turbines, for example, based on relative locations of turbines close together, on a thrust coefficient of each wind turbine 102 and / or on wind conditions predominant. Specifically, the Jensen Model can predict a speed deficit in the downstream wind turbine, for example, based on a wind speed, a wind direction and a buoyancy coefficient corresponding to the upstream wind turbine and a location of the wind turbine. downstream wind turbine. The speed deficit represents the whirlpool interactions between two wind turbines and then helps to identify the reference series of interacting wind turbines.
    [047] Furthermore, in step 206, one or more historical eddy models for the reference series of wind turbines in interaction are determined based on the historical values corresponding to the reference series. As used herein, the terms "series" or "series" are used to refer to a group of two or more quantities. Particularly, in one embodiment, each series of wind turbines corresponds to a pair of wind turbines. In other embodiments, however, each series may include three or more wind turbines. It can be seen that the whirlpool effects occur in series between the corresponding upstream and downstream turbine series interacting for a particular wind direction. A series estimation of the eddy effects is then aggregated to provide an estimate of the eddy effect at the park level. This serial estimation is determined using one or more historical whirlpool models generated for each reference series.
    23/41 [048] Consequently, in one realization, the historical values corresponding to each reference series are segregated from the historical values received to determine the historical whirlpool models. In one embodiment, the reference series can be identified using the engineering whirlpool models. In addition, for each of the reference series, segregated historical values provide, for example, predetermined values for different combinations of eddy parameters, such as wind direction, wind speed in upstream and downstream wind turbines, the angle of pitch, yaw misalignment and / or upstream wind turbine peak speed ratio.
    [049] Although several eddy parameters can be monitored simultaneously in a wind farm, in the present disclosure, different subsets of eddy parameters can be selected for different operating conditions, for example, during the day or night, during calm conditions or stormy weather, and / or to optimize different performance targets. In an exemplary implementation that aims to maximize a power output at the park level, the sub-series of eddy parameters includes values corresponding to a pitch angle, a peak speed ratio and a wind speed corresponding to the wind turbine at upstream and wind speed in a corresponding segregated downstream turbine for each reference series. In one embodiment, wind speeds can be measured directly or estimated from the turbine power, rotor speed and pitch angle measurements. In certain embodiments, the segregated values are processed to filter out noise data, for example, when the peak speed ratio is above or below the thresholds designated to provide more accurate modeling of the historical values of the eddy parameters.
    24/41 [050] In addition, in an accomplishment presently contemplated, the park control subsystem is adapted to a regression model through the corresponding segregated values for each reference series to determine the eddy models in historical series for different combinations of swirl parameters. In certain embodiments, the regression model employs machine learning to determine a statistical relationship between values of one or more selected eddy parameters and a predominant wind speed and direction in the upstream and downstream wind turbines in each reference series.
    [051] Particularly, in one example, the park control subsystem fits the regression model through the corresponding segregated values for each reference series using equation (1) = f (fi> s, TSRà ma , 0 cjma ><P drna others) 'ama (1) where ^ low ° corresponds to the wind speed in the downstream wind turbine, ^ up corresponds to the wind speed in the upstream wind turbine, β corresponds to a relative wind direction, s corresponds to a relative distance between the upstream and downstream wind turbines in each TSR rim series.
    of interacting wind turbines, ama corresponds to the speed ratio of the upstream wind turbine, 17 upwards corresponds to the pitch angle φ.
    of the upstream wind turbine, and above. corresponds to a misalignment of the upstream wind turbine yaw.
    [052] In one embodiment, equation (1) defines a transfer function that allows the prediction of whirlpool interactions in series as a ratio of wind speed in the downstream wind turbine to the wind speed in the upstream wind turbine. . Specifically, the regression model estimates the ratio as a function of the upstream wind direction β, the relative distance s between the upstream and downstream wind turbines, the
    25/41 upstream turbine tip speed ratio love, the angle θ Ψ step 17 above the wind turbine and upstream of the yaw misalignment of the wind turbine loves upstream.
    [053] In certain embodiments, the regression model can undergo machine learning to fine-tune the transfer function in order to provide an accurate forecast of the wind speed ratio in downstream and upstream wind turbines. The park control subsystem can then use a power model (turbine power as a function of wind speed and control settings) to predict the park-level power output of wind speeds predicted by historic whirlpool models . In one embodiment, the steps corresponding to receiving the historical values, identifying the reference series and / or determining the historical eddy models can be performed in an inactive mode, although the remaining steps illustrated in Figure 2 can be performed in real time. In an alternative, however, all steps of the method illustrated in Figure 2 can be performed in real time.
    [054] In one embodiment, the historical values of the eddy parameters, the historic eddy models and / or the predicted wind speed ratios can be stored in a lookup table in an associated storage repository. In addition, in step 208, historical eddy models are received, for example, in the park control subsystem for use in subsequent optimizations. In particular, in one embodiment, the park control subsystem receives the historic whirlpool models of wind turbines, turbine controllers and / or the storage repository.
    [055] Additionally, in step 210, the new values corresponding to at least some of the whirlpool parameters for wind turbines in the wind farm are received. As verified
    26/41 previously, environmental conditions in the wind farm, such as wind speed and direction, tend to vary continuously throughout the day. Thus, the swirling effects suffered by the wind farm can also vary throughout the day, thus meriting continuous adjustment of control settings. The continuous adjustment of the control settings, in turn, requires further evaluation of new values of the eddy parameters.
    [056] Consequently, in certain embodiments, new values corresponding, for example, to wind speed and direction, can be received in the park control subsystem at designated time intervals, for example, every ten minutes. Alternatively, new values can be received at random intervals, upon receipt of a user request, by determining a significant change (for example, a change greater than 5%) in consecutive measurements of the values of the eddy parameters and / or if a value of at least one eddy parameter is outside a specified threshold. In one embodiment, for example, the new values can be received if a change greater than 0.5 meters / second in the wind speed or greater than 5 degrees in the wind direction is observed.
    [057] Furthermore, in step 212, the new series of interacting wind turbines are identified from the wind turbines based on the new values. In one embodiment, the new series of interacting wind turbines can be defined for a particular wind direction using the method described above with reference to step 204.
    [058] Additionally, in step 214, a forecast swirl model at the park level can be developed for the new series of wind turbines in interaction based on one or more historic series swirl models and the new values. In one realization, the
    27/41 new values corresponding to each new series of interacting wind turbines are segregated. In addition, the park control subsystem develops a forecast progression model based on the segregated values corresponding to each new series and the historic eddy models. In one embodiment, the park control subsystem adapts the transfer functions corresponding to the historic eddy models to develop the park level forecasting model based on the new values. Alternatively, the park control subsystem aggregates historical eddy models to develop the park level forecasting model based on the new values. The prediction model at the park level, then developed, can be used to provide a robust estimate of eddy interactions, for example, a predicted prevailing wind speed ratio upstream to downstream for use in determining appropriate optimizations in the operation of the wind farm.
    [059] In step 216, one or more control settings for at least the new series of interacting wind turbines are adjusted based on the forecast swirl model at the park level. In particular, the forecasting whirlpool model at the park level can provide a forecast of how a change in certain control settings for upstream wind turbines in each new series can affect the operational efficiency of at least the corresponding downstream wind turbines . Consequently, the forecast swirl model at the park level can be used to determine the control settings for each wind turbine so that a desired park level performance target can be achieved and / or maintained.
    [060] However, the simultaneous determination of control settings for every wind turbine in the wind farm is a process
    28/41 computationally intensive and complex that can be difficult to implement in real time due to the large number of parameters that need to be evaluated. Consequently, in an accomplishment presently contemplated, the park control subsystem determines, sequentially and progressively, one or more control definitions for each wind turbine, at least in the new series of interacting wind turbines to optimize performance targets at the level the park.
    [061] Figure 3, for example, illustrates a schematic representation 300 that depicts an exemplary sequence 302 for determining the ideal control settings for wind turbines in aerodynamically interacting 1i-Tn in a wind farm. Specifically, Figure 3 illustrates the plurality of Ti-Tn wind turbines that can be classified into multiple series 308 to 324 of wind turbines in interaction based on a predominant wind direction 304 and a geometric layout of the wind farm. The series 308 to 324 can be identified so that each series includes at least one upstream wind turbine that activates at least one eddy downstream wind turbine, as previously described with reference to steps 204 and 212 of Figure 2.
    [062] In certain embodiments, the series of interacting wind turbines 308 to 324, for example, can correspond to the new series identified in step 212 of Figure 2 and can be represented in a structure similar to a sparse tree. Additionally, in an exemplary deployment, the park control subsystem is configured to sequentially determine the ideal control settings in a bottom-up approach so that the ideal control settings for a downstream wind turbine are determined in sequence of the definitions control systems for an upstream wind turbine in each of the new series.
    29/41 [063] For example, during the evaluation of the 308 series, the park control subsystem determines the appropriate control settings for the wind turbine further downstream Tu so that an individual power output [J (1) = P (1)] of Tu is maximized. In one embodiment, it can be assumed that the wind speed observed in Tu is equal to the free current wind speed (or at an arbitrary fixed value) when determining the power output for a generated control definition. Since the downstream wind turbine Tu is positioned at the bottom of the wind farm in view of the particular wind direction 304, the downstream wind turbine Tu suffers significant swirl effects. Consequently, the initiation of the optimization sequence 302 in the Tu wind turbine allows a substantial improvement in the power output P (1) of the Tu wind turbine without having to be responsible for the operation of the upstream Ti 0 turbine.
    [064] Subsequently, the park control subsystem determines the appropriate control settings for the upstream wind turbine T 10 in order to maximize the combined power output [J (2) = P (2) + J (1)] produced by the Tio and Tu turbines in view of the forecast swirl model at the park level. In one embodiment, the control settings can be determined based on an assumption that the wind speed observed in Tw is equal to the speed of the free current wind, although the wind speed in Th is determined in view of the model of forecast swirl at park level. Furthermore, to maximize J (2), the park control subsystem determines the effect of the control settings corresponding to the upstream wind turbine T í0 on the power output produced by the downstream turbine Tu based on the eddy model at the park level and in the control settings that were previously determined for the downstream wind turbine Tn · [065] Additionally, for the 310 series that includes the turbines
    30/41 wind turbines T w and T 4 , the control settings for the upstream wind turbine T 4 can be determined in order to maximize the combined power output [J (3j = P (3) + J (2)] for the upstream turbine T 4 , and for the corresponding downstream turbines Tw, and Tu. In particular, the control settings for the upstream turbine T 4 can be determined to maximize the power output [Jf3) already assuming that the speed of the wind observed in T 4 is equal to the speed of the free current wind (or to an arbitrary fixed value), and using the whirlpool model at the park level and the control settings previously determined for the Ti 0 and Tu turbines as restrictions. Similarly, the control definitions for the series of wind turbines positioned along other branches of the sparse tree structure can be determined so that, at each positional level, the combined power output of the wind turbine at the level and levels previous ones is maximized, in turn, maximizing the power output at park level.
    [066] Occasionally, the control settings determined for certain wind turbines in the wind farm may cause one or more wind turbine performance parameters to be outside the allowable limits specified for a wind speed observed in wind turbines. For example, a given peak speed ratio and pitch angle combination for a particular wind turbine can result in a power output that is greater than a nominal limit for the wind speed observed in the wind turbine. In this situation, there may be a need to adjust the control settings, for example, to reduce the peak speed ratio and / or to increase the pitch angle until the power output is equal to the nominal limit. Consequently, once the appropriate control settings for wind turbines are determined, as described with reference to Figure 3, the
    31/41 control in a subset of wind turbines can be readjusted in relation to the wind speeds expected in these wind turbines.
    [067] In one embodiment, this readjustment of control settings for the wind turbine subseries can be performed in a top-down manner. For example, the control settings for the wind turbine further up in the sparse tree structure can be adjusted, first, followed by the control settings for the wind turbines at the subsequent downstream level. Top-down adjustment allows for more efficient computations, as an expected wind speed for each selected wind turbine can be calculated based on the wind speed measured at the upstream turbine in the sparse tree structure and corresponding control settings for all corresponding upstream turbines. The calculated wind speed can, in turn, be used to readjust the control settings so that the performance of the selected wind turbine remains within the allowable limits.
    [068] This sequential determination and / or readjustment of control settings for wind turbines allows for a serial resolution of the optimization problem, thereby reducing the complexity and computational effort associated with optimizing a performance target at the level of general park. In certain embodiments, the park-level optimization described with reference to Figures 2 to 3 can be implemented continuously to ensure that the desired performance goals remain within the designated limits.
    [069] Additionally, in one embodiment, the present method can be implemented in a delayed optimization mode, in which the ideal stored control definitions that were previously determined for the historical values of selected swirl parameter combinations cannot be calculated continuously however
    32/41 can be used to adjust wind turbine operations in real time. In certain embodiments, the stored control definitions may be updated periodically in view of varying environmental conditions at designated time intervals, or when the eddy parameter values are outside the corresponding designated thresholds.
    [070] In an alternative, however, model updates can be performed in real time to allow a more accurate estimation of the prevailing whirlpool conditions, which in turn provide more precise adjustments to the control settings for each wind turbine. The certain exemplary realizations of delayed and real-time optimization of the wind farm operation will be described in more detail with reference to Figures 4 to 5.
    [071] In particular, Figure 4 illustrates a flowchart 400 that depicts an exemplary method for optimizing the operation of a wind farm in a delayed optimization mode. The method starts at step 402, in which the environmental information and, optionally, the operational information corresponding to a wind farm are received to be used in combination with the geometric information. Additionally, in step 404, the forecast swirl models at the park level and / or the historical swirl models corresponding to the reference series of wind turbines are received.
    [072] In addition, in step 406, the different series of interacting wind turbines can be identified from the wind turbines for one or more selected combinations of eddy parameters. In one embodiment, selected combinations of the eddy parameters include the selected values of upstream wind speeds, downstream wind speeds and selected wind directions. Wind speeds and directions, for example, can be selected from
    33/41 from stored historical climate information that includes frequency distributions in the wind direction batch of average wind speeds for the wind farm.
    [073] In step 408, for each combination of selected combinations of eddy parameters, the ideal control settings are determined for the different series of wind turbines in interaction based on the historic eddy models and / or the forecast eddy models at park level. In addition, in step 410, the optimal control definitions for the different definitions are stored as a function of the corresponding selected combination of swirl parameters in a storage repository. Alternatively, in one embodiment, the method described with reference to Figure 2 is performed or simulated instead of steps 402 to 408 so that different combinations of eddy parameters determine the appropriate control settings for the wind turbines in the wind farm. The resulting control definitions determined over time can be recorded in memory to generate an appropriate lookup table that correlates the control definitions to selected combinations of eddy parameters. In particular, once steps 402 to 410 have been contemplated, the lookup table can be used to adjust the control settings in response to the recently obtained values of the eddy parameters.
    [074] In step 412, new values corresponding to at least some of the eddy parameters are received. In one realization, the new values relate to wind speed, wind direction and / or operational information for each of the wind turbines. The wind direction, in one example, corresponds to a median wind direction across all wind turbines and can be determined from yaw positions and / or using a pinwheel. Furthermore, in this
    34/41 example, the wind speed for the determined wind direction corresponds to a median wind speed in the upstream wind turbines. In certain embodiments, the wind speed can be estimated based on the power, the rotor speeds and / or the pitch of the wind turbines. Alternatively, the wind speed can be estimated using an anemometer.
    [075] In addition, in step 414, one or more control definitions for the different series of wind turbines can be interpolated from the ideal stored control definitions and the new values of the eddy parameters. In one embodiment, the new values, for example, the wind speed and direction received in step 412, can be made compatible with the stored values of a selected combination of swirl parameters. In certain embodiments, a stored correlation, for example, the lookup table can be consulted to identify the ideal stored definitions as a function of the wind speed and direction values received in step 412. If the lookup table does not include values exact wind speed and direction received in step 412, the park control subsystem can be configured to interpolate the ideal settings for each wind turbine from the ideal stored control settings corresponding to the closest wind speed and values .
    [076] In step 416, the different series of interacting wind turbines are operated using interpolated values from the control definitions. In certain realizations, the interpolated values of the control definitions can be stored in the storage repository for subsequent optimizations. Additionally, in certain additional realizations, the ideal control settings originally stored in the storage repository can be updated occasionally with
    35/41 based on interpolated values to be responsible for variations in the turbine and / or performance values at the park level.
    [077] Such inactive optimization of the look-up table can be used during the operation of the wind farm to enable updates of the control definitions in view of the constantly varying environmental conditions and limited instrumentation. However, the predetermined optimized lookup table approach can provide a limited improvement only in the performance of wind farms that have significant variations in the park's terrain, frequently changing wind speeds and directions and / or frequent occurrences of one or more downtime. more among wind turbines.
    [078] Figure 5 illustrates a flow chart 500 that depicts an exemplary method for optimizing the operation of a wind farm, in which the underlying swirl models are adjusted and real time. As used in this document, the term real-time can be used to refer to a time delay of about an hour from collecting the wind farm's operational information to adjusting the eddy models used to determine the control settings ideal for wind turbines. The method starts at step 502, in which the environmental information and optional operational information corresponding to a wind farm are received for use in combination with the geographical information. In one embodiment, the environmental information includes a wind speed and direction detected in different wind turbines, while the operational information corresponds to the states in operation and / or out of operation of the wind turbines in the wind farm.
    [079] In addition, in step 504, one or more historical eddy models corresponding to the reference series of wind turbines can be received. In one embodiment, the swirl models
    36/41 can be determined using the method described with reference to step 206 in Figure 2.
    [080] Additionally, in step 506, new values corresponding to at least some of the eddy parameters are received. Typically, environmental conditions at the wind farm, for example, wind speed and direction, tend to vary continuously throughout the day. The whirlpool effects suffered by the wind farm can then also vary throughout the day, thus deserving a continuous update of the whirlwind model. Consequently, in certain embodiments, the new wind speed values, wind direction and states in operation and / or out of operation for wind turbines can be received in the park control subsystem at designated time intervals.
    [081] In general, wind conditions and permanent turbine performance in a wind farm are known to vary in time intervals of approximately ten minutes. Therefore, in one realization, the park control subsystem can request that the new values be delivered every ten minutes. Alternatively, the new values can be received at random intervals, upon receipt of a user request, the determination of a significant change in consecutive measurements and / or if one of at least one eddy parameters is outside a designated threshold.
    [082] Additionally, in step 508, the new series of interacting wind turbines are intended from the plurality of wind turbines based on the new values and operational information. In one embodiment, the new series can be identified using the method described with reference to steps 204 and 212 in Figure 2. In addition, the operational status of each wind turbine in real time is considered
    37/41 to identify the new series. Thus, if a wind turbine is not in operation during a particular optimization period, the wind turbine will not contribute to the whirlwind effect and therefore will not be considered while identifying the new series. However, this wind turbine can be considered during another optimization period when the wind turbine is in active operation.
    [083] Additionally, in step 510, a forecast swirl model at the park level is developed for the new series of wind turbines in interaction based on one or more historical swirl models, new values and operational information. In one embodiment, the forecast swirl model at the park level is developed using the method described with reference to step 214 in Figure 2. In certain embodiments, the development of the forecast swirl models at the park level requires updating a swirl model at the park level determined previously determined based on the new values and operational information. The whirlpool model at the park level, then developed and / or updated, can then be used to predict the desired performance parameter values, for example, the power output and / or the fatigue loads suffered by different turbines in the wind farm.
    [084] However, frequent changes in environmental conditions, for example, a sudden change in speed in the direction of the wind or, occasionally, can make the whirlpool model at the park level developed in step 510 less relevant. Consequently, in step 512, it can be determined whether a revision capacity of the whirlpool model at the park level developed in step 510 is satisfactory. For this purpose, the desired performance parameters for wind turbines in the wind farm can be measured using sensors,
    38/41 for example, the monitoring devices 110 of Figure 1. In addition, the measured values can be compared to the values of the desired performance parameters predicted by the whirlpool model at the park level.
    [085] In one embodiment, if the expected values of the desired performance parameters differ by more than a designated amount (for example, more or less than 5%) from the corresponding measured values, the forecasting capacity of the eddy model at the level of park can be determined to be unsatisfactory. Consequently, the control can pass to cover 506 for the subsequent steps of the method of Figure 5, which are repeated until the predicted values of the power output at the park level and / or the fatigue loads are substantially compatible with the measured values corresponding, thus signifying the precision of the whirlpool model at the park level. In certain embodiments, the method in Figure 5 can also be repeated after one or more designated time intervals to allow for continuous wind farm optimization operations in view of the frequent changes in eddy-causing environmental conditions.
    [086] When a difference between the predicted and measured values of the desired performance parameters is determined to be less than the designated quantity, in step 514, one or more control definitions for at least the new series of interacting wind turbines are adjusted based on the park level forecast swirl model developed in step 510. In particular, park level forecast swirl models can be used to determine the control settings for each wind turbine so that a performance target desired park level can be achieved. In one embodiment, for example, the control definitions for one or more of the
    39/41 interacting wind turbines are adjusted sequentially and in pairs so that the power output at the park level is maximized.
    [087] The achievements of the present disclosure then present a data driven swirl modeling approach that uses the real-time values of swirl parameters to generate robust forecast swirl models at park level. In a more specific realization, the data-driven approach uses the eddy parameters assembled to aerodynamically identify the series of interacting wind turbines and estimate the corresponding eddy interactions (in series). Serial swirl interactions, in turn, are used to generate more accurate real-time forecast swirl models at the park level.
    [088] In an exemplary deployment, the use of the current data-driven swirl modeling approach resulted in a data correlation greater than 60% compared to conventional engineering swirl models. Figure 6, for example, illustrates a graphical representation 600 that depicts a comparison of energy gains achieved using a wind farm baseline operation, a classic wind farm model (an engineering swirl model) and an embodiment of the present method described with reference to Figures 2 to 5. The data-driven eddy modeling approach is responsible for continuous variations in eddy-causing environmental conditions, for example, speed, direction, intensity and / or wind turbulence that physics-based engineering whirlpool models cannot capture precisely.
    [089] Consequently, as evident from the portraits in Figure 6, the present method provides an energy gain greater than 602 than the gain achieved using the 604 engineering swirl models
    40/41 and / or 606 baseline operations. In particular, the use of prevailing environmental conditions and operating states of individual turbines allows the capture of whirlpool interactions to be suffered in real time by downstream wind turbines, thereby , allowing the determination of a more accurate forecast swirl model at park level. Additionally, the determination of the forecast swirl models at the park level through the series evaluation of the swirl interactions reduces the computational effort, thereby allowing a faster optimization of one or more performance targets selected for the wind farm. Specifically, forecasting whirlpool models at the park level assist in determining the optimal control settings for the different wind turbines in the wind farm in order to improve overall performance targets.
    [090] It should be noted that the examples, the aforementioned demonstrations and the process steps that can be performed by certain components of the present systems, for example, by the turbine controllers 106 and / or by the park control subsystem 108 in Figure 1 can be deployed by suitable code on a processor-based system. For this purpose, the processor-based system, for example, may include a general-purpose or specific-purpose computer. It should also be noted that the different deployments of the present disclosure may carry out some or all of the steps described in this document in different orders or in a substantially concurrent manner.
    [091] Additionally, functions can be implemented in a variety of programming languages, including, but not limited to, Ruby, Hypertext Preprocessor (PHP), Perl, Delphi, Python, C, C ++ or Java. Such code can be stored or adapted for storage in
    41/41 one or more machine-readable tangible media, for example, on data repository chips, local or remote hard drives, optical discs (ie CDs or DVDs), solid state drives, or other accessible media by the processor-based system to execute the stored code.
    [092] Although the specific features of the achievements of the present disclosure may be shown and / or described in relation to some drawings and not in others, this is only for convenience. It should be understood that the resources, structures and / or features described can be combined and / or used interchangeably in any suitable manner in the various realizations, for example, to build additional sets and methods for use in wind farm optimization.
    [093] Although only certain features of the present disclosure have been illustrated and described in this document, many modifications and changes will occur for those skilled in the art. Therefore, it should be understood that the appended claims are intended to cover all such modifications and changes, as they fall within the true spirit of the invention.
    1/6
    权利要求:
    Claims (20)
    [1]
    Claims
    1. METHOD TO OPTIMIZE THE OPERATION OF A WIND FARM, characterized by the fact that it comprises:
    - receive new values corresponding to at least some whirlpool parameters for wind turbines in the wind farm;
    - identify new series of wind turbines in interaction from the wind turbines based on the new values;
    - develop a forecast swirl model at the park level for the new series of wind turbines in interaction based on the new values and historical swirl models determined using the historical values of the swirl parameters corresponding to the reference series of wind turbines in interaction at the wind farm; and
    - adjust one or more control settings for at least the new series of wind turbines in interaction based on the forecast swirl model at the park level.
    [2]
    2. METHOD, according to claim 1, characterized by the fact that it additionally comprises:
    - receive the historical values of the whirlpool parameters corresponding to the wind turbines before receiving the new values;
    - identify the reference series of wind turbines in interaction from the wind turbines based on historical values; and
    - determine the one or more historical eddy models for the reference series of wind turbines in interaction based on the historical values corresponding to the reference series of wind turbines in interaction.
    [3]
    3. METHOD, according to claim 2, characterized by the fact that determining the historical eddy models comprises adapting the historical values corresponding to each of the series
    2/6 references of wind turbines interacting with the use of a regression-based model.
    [4]
    4. METHOD, according to claim 2, characterized by the fact that determining historical eddy models comprises computing a ratio of downstream wind speed to upstream wind speed as a function of wind direction in a wind turbine at upstream, relative upstream and downstream wind turbine locations and the one or more control definitions corresponding to the upstream wind turbine using the regression-based model.
    [5]
    5. METHOD, according to claim 1, characterized by the fact that identifying the new series of interacting wind turbines comprises using at least one subseries of the new values and the geometric layout of the wind farm.
    [6]
    6. METHOD, according to claim 1, characterized by the fact that the receipt, identification, development and adjustment are carried out in one or more designated time intervals.
    [7]
    7. METHOD, according to claim 1, characterized by the fact that it additionally comprises:
    - continuously monitor the whirlpool parameters for wind turbines; and
    - repeat receipt, identification, development and adjustment when a change in a monitored value of one or more among the eddy parameters is outside a corresponding threshold.
    [8]
    8. METHOD according to claim 1, characterized by the fact that different historical eddy patterns are determined for different combinations of eddy parameters.
    [9]
    9. METHOD, according to claim 1, characterized by the fact that the eddy parameters include wind direction,
    3/6 wind speed in an upstream wind turbine, wind speed in a downstream wind turbine, wind turbulence, wind shear, wind rotation, ambient temperature, pressure, humidity or combinations thereof.
    [10]
    10. METHOD, according to claim 1, characterized by the fact that the eddy parameters comprise at least one of a peak speed ratio, a pitch angle, a yaw misalignment and an operational state of each of the turbines wind farms.
    [11]
    11. METHOD, according to claim 1, characterized by the fact that the eddy parameters comprise information of the geometric layout of the wind farm.
    [12]
    12. METHOD, according to claim 1, characterized by the fact that adjusting the control settings comprises sequentially determining the control settings for a downstream wind turbine followed by an upstream wind turbine in each of the new series of wind turbines interacting to achieve one or more desired performance goals.
    [13]
    13. METHOD, according to claim 1, characterized by the fact that adjusting the control definitions comprises sequentially determining the control definitions for each of the new series of interacting wind turbines positioned in the wind farm in a sparse tree structure of so that, at each level, position in the sparse tree structure, a combined power output of the wind turbines at the positional level and at the previous positional levels in the sparse tree structure is maximized.
    [14]
    14. METHOD, according to claim 13, characterized by the fact that it additionally comprises readjusting the definitions of
    4/6 control for a subset of wind turbines, if the control settings determined for the subset of wind turbines result in a performance parameter that is outside a specified permissible limit for an expected wind speed in the subset of wind turbines, in that readjusting the control settings comprises sequentially determining the control settings for each subset of wind turbines in a top-down manner.
    [15]
    15. METHOD, according to claim 1, characterized by the fact that achieving the desired performance targets comprises reducing fatigue loads on wind turbines in the new series of wind turbines in interaction below a first threshold, increasing annual energy production wind farm above a second threshold or a combination thereof.
    [16]
    16. METHOD, according to claim 1, characterized by the fact that each of the series of interacting wind turbines comprises a pair of wind turbines.
    [17]
    17. METHOD FOR OPERATING A WIND FARM, characterized by the fact that it comprises:
    - assemble historical eddy models for different series of interacting wind turbines in the wind farm based on historical values of selected combinations of eddy parameters corresponding to the series of interacting wind turbines;
    - determine optimal control settings for each wind turbine in the series of interacting wind turbines for each of the selected combinations of eddy parameters based on the historic eddy model;
    - store the ideal control settings for each wind turbine as a function of the selected combination of
    5/6 swirl;
    - receive new values of the eddy parameters obtained in a subsequent period of time after obtaining the historical values; and
    - determine the control settings for wind turbines in each of the new series of wind turbines using the new values and stored control settings.
    [18]
    18. SYSTEM TO OPTIMIZE THE OPERATION OF A WIND FARM, characterized by the fact that it comprises
    - a plurality of wind turbines;
    - one or more monitoring devices configured to measure values from a plurality of eddy parameters for one or more among the plurality of wind turbines; and
    - a park control subsystem operationally coupled, at least, to the monitoring devices and programmed to:
    - receive new values corresponding to at least some whirlpool parameters for wind turbines in the wind farm;
    - identify new series of wind turbines in interaction from the plurality of wind turbines based on the new values;
    - develop a forecast swirl model at the park level for the new series of wind turbines in interaction based on the new values and historical swirl models determined using the historical values of the swirl parameters corresponding to the reference series of wind turbines in interaction at the wind farm; and
    - adjust one or more control settings for at least the new series of wind turbines in interaction based on the forecast swirl model at the park level.
    6/6
    [19]
    19. SYSTEM, according to claim 18, characterized by the fact that the monitoring devices comprise rotor speed encoders, pitch angle encoders, electrical power transducers, anemometers, pinwheels, yaw position encoders, or combinations thereof.
    [20]
    20. SYSTEM, according to claim 18, characterized by the fact that the park control subsystem comprises a centralized processing system.
    1/6
    类似技术:
    公开号 | 公开日 | 专利标题
    BR102015009550A2|2018-03-06|METHOD AND SYSTEM FOR OPTIMIZING OPERATION OF A WIND PARK AND METHOD FOR OPERATING A WIND PARK
    EP3023636B1|2021-04-14|Systems and methods for optimizing operation of a wind farm
    US9644612B2|2017-05-09|Systems and methods for validating wind farm performance measurements
    CN109154281B|2021-09-24|System and method for predicting power output of a wind farm
    EP2581600B1|2018-12-12|Method and system for control of wind turbines
    US9797377B2|2017-10-24|System and method for controlling a wind farm
    US10247170B2|2019-04-02|System and method for controlling a dynamic system
    ES2681024T3|2018-09-11|Wind turbine and procedure for the control of a wind turbine or a wind farm
    US10253758B2|2019-04-09|System and method for optimizing wind farm performance
    US10605228B2|2020-03-31|Method for controlling operation of a wind turbine
    US10385829B2|2019-08-20|System and method for validating optimization of a wind farm
    CN109488528B|2020-03-17|Method and device for adjusting fan yaw system
    WO2018059259A1|2018-04-05|Method and system of yaw control of wind turbines in a wind turbine farm
    CN113874626A|2021-12-31|System and method for evaluating and verifying wind turbine and wind farm performance
    EP3009670B1|2018-01-03|Method to control the operation of a wind energy installation, including generating a 3D wind speed model, and wind energy installation
    EP3608538A1|2020-02-12|Model-based repowering solutions for wind turbines
    Ahmad et al.2015|Analysis of two onshore wind farms with a dynamic farm controller.
    Bao et al.2019|Identification of yaw error inherent misalignment for wind turbine based on SCADA data: A data mining approach
    Kanev et al.2016|On the application of LiDARs in wind farm control
    DuPont et al.2014|Employing wind farm performance data for model validation and turbine layout/geometry optimization
    Dasari2018|Near-wake studies of a utility-scale wind turbine using natural snowfall-based visualization and Particle Image Velocimetry
    同族专利:
    公开号 | 公开日
    DK2940296T3|2017-08-28|
    ES2636659T3|2017-10-06|
    US20150308416A1|2015-10-29|
    EP2940296A1|2015-11-04|
    EP2940296B1|2017-06-14|
    CN105041572B|2019-05-03|
    CA2888737A1|2015-10-29|
    CN105041572A|2015-11-11|
    US9551322B2|2017-01-24|
    CA2888737C|2018-09-04|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    BR9915956B1|1998-12-04|2011-10-18|pressure sensor, and method for sensing pressure.|
    DE19948196A1|1999-10-06|2001-05-17|Aloys Wobben|Process for operating a wind farm|
    DE10137272A1|2001-07-31|2003-02-27|Aloys Wobben|Early warning system for wind turbines|
    NL1021078C1|2002-07-15|2004-01-16|Energieonderzoek Ct Petten Ecn|Method and device concerning flow energy such as a wind farm.|
    NL1023666C2|2003-06-14|2004-12-20|Energieonderzoek Ct Petten Ecn|Method or device for extracting energy from a flowing fluid.|
    DE102004056254B4|2004-11-22|2006-11-09|Repower Systems Ag|Method for optimizing the operation of wind turbines|
    US20070124025A1|2005-11-29|2007-05-31|General Electric Company|Windpark turbine control system and method for wind condition estimation and performance optimization|
    DK200700630A|2007-04-27|2008-05-10|Lm Glasfiber As|Design of group of wind power plants|
    DE102007022705A1|2007-05-15|2008-11-20|Siemens Ag|Method for operating a wind farm with a plurality of wind turbines|
    DE102007036444A1|2007-08-02|2009-02-05|Nordex Energy Gmbh|Wind farm with a large number of wind turbines and procedures for operating the wind farm|
    US20090099702A1|2007-10-16|2009-04-16|General Electric Company|System and method for optimizing wake interaction between wind turbines|
    US8364323B2|2007-12-14|2013-01-29|Mitsubishi Heavy Industries, Ltd.|Wind power generation system and operation control method thereof|
    US20090299780A1|2008-05-29|2009-12-03|Abhinanda Sarkar|Method and apparatus for determining and/or providing power output information of wind turbine farms|
    DE102008039429A1|2008-08-23|2010-02-25|DeWind, Inc. , Irvine|Method for controlling a wind farm|
    BRPI0913564B1|2008-09-19|2020-12-08|Wobben Properties Gmbh|fast-run wind turbine|
    US7941304B2|2009-04-30|2011-05-10|General Electric Company|Method for enhancement of a wind plant layout with multiple wind turbines|
    EP2354541B1|2010-01-20|2014-09-17|Siemens Aktiengesellschaft|Wind farm power control based on matrix reflecting a power load distribution between individual wind turbines|
    CN101852172A|2010-03-09|2010-10-06|山东科技大学|Method for calculating input wind speed of wind generating sets according to wake effect in wind power station|
    GB2481461A|2010-06-21|2011-12-28|Vestas Wind Sys As|Control of a downstream wind turbine in a wind park by sensing the wake turbulence of an upstream turbine|
    DE102010026244A1|2010-07-03|2012-01-05|Joachim Falkenhagen|Method for controlling e.g. rotor rotation of wind turbine in large wind farm, involves programming appropriate control system such that renouncement of yielding potential is performed as function of wind direction and position of turbine|
    US8035241B2|2010-07-09|2011-10-11|General Electric Company|Wind turbine, control system, and method for optimizing wind turbine power production|
    CN101949363A|2010-09-21|2011-01-19|山东科技大学|Method for grouping wind generating sets by taking input wind speed and random fluctuation of wind direction of wind farm into consideration|
    CN102012956B|2010-11-30|2016-01-20|山东科技大学|A kind of wind energy turbine set equivalence method considering the random fluctuation of wind energy turbine set input wind speed and direction|
    US20110193344A1|2010-12-29|2011-08-11|Vestas Wind Systems A/S|Control Network for Wind Turbine Park|
    US8606418B1|2011-03-18|2013-12-10|Rockwell Collins, Inc.|Wind prediction for wind farms through the use of weather radar|
    CN102235313B|2011-06-30|2013-01-16|内蒙古电力勘测设计院|Regular arrangement optimization method of fans in flat terrain|
    CN102289539B|2011-06-30|2013-05-01|内蒙古电力勘测设计院|Method for optimizing fan layout for improving utilization rate of wind energy|
    CN102270256B|2011-06-30|2013-04-03|内蒙古电力勘测设计院|Wind field planning method for flat region|
    EP2557311A1|2011-08-12|2013-02-13|kk-electronic a/s|A method for controlling a wind power park and a wind power park controlled by such method|
    DE102011081241A1|2011-08-19|2013-02-21|Repower Systems Se|Energy yield loss determination of a wind turbine|
    EP2758660A1|2011-09-13|2014-07-30|Vestas Wind Systems A/S|A method for improving large array wind park power performance through active wake manipulation reducing shadow effects|
    US9644610B2|2011-12-06|2017-05-09|Vestas Wind Systems A/S|Warning a wind turbine generator in a wind park of an extreme wind event|
    EP2604853A1|2011-12-15|2013-06-19|Siemens Aktiengesellschaft|Method of controlling a wind turbine|
    US20120133138A1|2011-12-22|2012-05-31|Vestas Wind Systems A/S|Plant power optimization|
    US9201410B2|2011-12-23|2015-12-01|General Electric Company|Methods and systems for optimizing farm-level metrics in a wind farm|
    WO2013110327A1|2012-01-25|2013-08-01|Abb Research Ltd|Wind park with real time wind speed measurements|
    EP2644889B1|2012-03-29|2015-05-13|ALSTOM Renewable Technologies|Detecting a wake situation in a wind farm|
    KR101337651B1|2012-05-04|2013-12-05|삼성중공업 주식회사|System and method for output control of wind farm|
    US20130300115A1|2012-05-08|2013-11-14|Johnson Controls Technology Company|Systems and methods for optimizing power generation in a wind farm turbine array|
    DE102012011210A1|2012-06-06|2013-12-12|Rwe Innogy Gmbh|Target value-dependent control of a wind farm|
    DK2877741T3|2012-07-27|2019-05-20|Univ Texas Tech System|System and method for evaluating wind flow fields using remote sensing devices|
    US9165092B2|2012-07-31|2015-10-20|International Business Machines Corporation|Wind farm layout in consideration of three-dimensional wake|
    US9617975B2|2012-08-06|2017-04-11|General Electric Company|Wind turbine yaw control|
    CN102913399A|2012-11-06|2013-02-06|中国科学院工程热物理研究所|Plasma flow control method for reducing the wake loss of a wind turbine|
    CN103020462B|2012-12-21|2016-01-13|华北电力大学|Take into account the wind energy turbine set probability output power calculation algorithms of complicated wake effect model|
    US9512820B2|2013-02-19|2016-12-06|Siemens Aktiengesellschaft|Method and system for improving wind farm power production efficiency|WO2015134655A2|2014-03-05|2015-09-11|Adeptdc Co.|Systems and methods for intelligent controls for optimal resource allocation for data center operations|
    ES2563092B1|2014-09-10|2016-12-19|Acciona Windpower, S.A.|Wind turbine control method|
    EP3037657A1|2014-12-23|2016-06-29|ABB Technology AG|Optimal wind farm operation|
    US20160215759A1|2015-01-28|2016-07-28|Alliance For Sustainable Energy, Llc|Methods and systems for wind plant power optimization|
    US10156224B2|2015-03-13|2018-12-18|General Electric Company|System and method for controlling a wind turbine|
    US9881108B2|2015-05-29|2018-01-30|One Energy Enterprises Llc|Method of evaluation wind flow based on conservation of momentum and variation in terrain|
    EP3121442B1|2015-07-20|2018-03-14|ALSTOM Renewable Technologies|Operating wind turbines|
    US20170038276A1|2015-08-04|2017-02-09|Solar Turbines Incorporated|Monitoring System for Turbomachinery|
    EP3380725B1|2015-11-26|2019-10-16|Vestas Wind Systems A/S|A method for monitoring and assessing power performance changes of a wind turbine|
    US10883475B2|2015-11-26|2021-01-05|Vestas Wind Systems A/S|Method for monitoring and assessing power performance changes of a wind turbine|
    DK178991B1|2015-12-22|2017-07-31|Envision EnergyCo Ltd|Method and system of operating a wind turbine farm|
    JP6405324B2|2016-01-29|2018-10-17|三菱重工業株式会社|Wind power generator and operation method thereof|
    DE102016103254A1|2016-02-24|2017-08-24|Wobben Properties Gmbh|Method for determining an equivalent wind speed|
    EP3423709B1|2016-03-03|2020-05-06|Vestas Wind Systems A/S|A method for determining a yaw position offset of a wind turbine|
    US10385829B2|2016-05-11|2019-08-20|General Electric Company|System and method for validating optimization of a wind farm|
    EP3464891B1|2016-06-07|2022-01-19|Vestas Wind Systems A/S|Adaptive control of a wind turbine by detecting a change in performance|
    US10247170B2|2016-06-07|2019-04-02|General Electric Company|System and method for controlling a dynamic system|
    US10247171B2|2016-06-14|2019-04-02|General Electric Company|System and method for coordinating wake and noise control systems of a wind farm|
    EP3299614B1|2016-09-26|2020-11-11|Siemens Gamesa Renewable Energy A/S|Method, device and system for managing a wind farm|
    CN107869420B|2016-09-27|2020-08-14|远景能源（江苏）有限公司|Method and system for controlling yaw of wind turbine farm|
    JP2018059450A|2016-10-06|2018-04-12|株式会社日立製作所|Wind power generation device and control method of wind power generation device|
    JP6869685B2|2016-10-06|2021-05-12|株式会社日立製作所|Wind farm and wind farm|
    CN108150360A|2016-12-05|2018-06-12|北京金风科创风电设备有限公司|The method and apparatus for detecting the equivalent load of Wind turbines|
    CN108256712B|2016-12-29|2021-03-16|北京金风科创风电设备有限公司|Control method and device for wind power plant group|
    US10598157B2|2017-02-07|2020-03-24|International Business Machines Corporation|Reducing curtailment of wind power generation|
    US10330081B2|2017-02-07|2019-06-25|International Business Machines Corporation|Reducing curtailment of wind power generation|
    JP2018135772A|2017-02-21|2018-08-30|株式会社日立製作所|Control device for plural wind power generators, and control method for wind farm or plural wind power generators|
    US11136961B2|2017-06-01|2021-10-05|General Electric Company|System and method for optimizing power output of a wind turbine during an operational constraint|
    EP3642480A1|2017-06-22|2020-04-29|General Electric Company|System and method for coordinating wake and noise control systems of a wind farm|
    DE102017009838A1|2017-10-23|2019-04-25|Senvion Gmbh|Control system and method for operating multiple wind turbines|
    DE102017009837A1|2017-10-23|2019-04-25|Senvion Gmbh|Control system and method for operating a plurality of wind turbines|
    CN107956638B|2017-11-15|2019-08-16|三一重能有限公司|Blower group of planes control method, control system and wind field|
    US10731630B2|2018-01-03|2020-08-04|General Electric Company|Extended reaction power for wind farms|
    WO2019141330A1|2018-01-22|2019-07-25|Vestas Wind Systems A/S|Control of a wind energy park comprising airborne wind energy systems|
    EP3517774A1|2018-01-25|2019-07-31|Siemens Gamesa Renewable Energy A/S|Method and apparatus for cooperative controlling wind turbines of a wind farm|
    EP3527817A1|2018-02-19|2019-08-21|Siemens Aktiengesellschaft|Methods and devices for automated determination and/or compensating for the influence of a wake vortex on a wind turbine|
    EP3533997A1|2018-02-28|2019-09-04|Siemens Gamesa Renewable Energy A/S|Estimating free-stream inflow at a wind turbine|
    DE102018003168A1|2018-04-18|2019-10-24|Senvion Gmbh|Method and device for operating a wind energy plant and wind energy plant|
    CN108708825B|2018-05-15|2019-11-26|浙江运达风电股份有限公司|A kind of wind power plant Yaw control method considering unit wake flow|
    CN108798997B|2018-06-28|2020-02-07|北京金风科创风电设备有限公司|Control method, device, controller and system of wind generating set|
    US20210310461A1|2018-07-31|2021-10-07|Alliance For Sustainable Energy, Llc|Distributed reinforcement learning and consensus control of energy systems|
    US11015576B2|2018-08-13|2021-05-25|Inventus Holdings, Llc|Wind turbine control system including an artificial intelligence ensemble engine|
    US10605228B2|2018-08-20|2020-03-31|General Electric Company|Method for controlling operation of a wind turbine|
    DK201870706A1|2018-10-31|2020-06-09|Vattenfall Ab|A dynamic optimisation strategy for improving the operation of a wind farm|
    CN111126651A|2018-10-31|2020-05-08|北京金风科创风电设备有限公司|Wind power plant noise prediction method, device and system|
    US10815972B2|2019-03-22|2020-10-27|General Electric Company|System and method for assessing and validating wind turbine and wind farm performance|
    WO2020249173A1|2019-06-14|2020-12-17|Vestas Wind Systems A/S|A method for controlling a wind farm under turbulent wind conditions|
    DE102019119774A1|2019-07-22|2021-01-28|fos4X GmbH|Method for controlling a wind park, control module for a wind park and wind park|
    US10954919B1|2019-10-02|2021-03-23|General Electric Company|Machine-learning model-based analytic for monitoring wind farm power performance|
    US11261846B2|2019-11-01|2022-03-01|General Electric Company|System and method for designing and operating a wind turbine power system based on statistical analysis of operational and/or grid data thereof|
    EP3839251A1|2019-12-20|2021-06-23|Wobben Properties GmbH|Wind farm optimisation|
    US20210246875A1|2020-02-06|2021-08-12|General Electric Company|System and method for optimizing wake management in wind farms|
    US11060504B1|2020-02-07|2021-07-13|General Electric Company|Systems and methods for continuous machine learning based control of wind turbines|
    WO2021190764A1|2020-03-27|2021-09-30|Winji Ag|Device, method and computer program for evaluating the operation of a power plant|
    CN111614088B|2020-06-09|2021-09-21|三一重能股份有限公司|Energy management method considering wake flow influence|
    CN111980857A|2020-08-26|2020-11-24|上海电气风电集团股份有限公司|Closed-loop control method and device for wind power plant and computer readable storage medium|
    US11231012B1|2020-09-22|2022-01-25|General Electric Renovables Espana, S.L.|Systems and methods for controlling a wind turbine|
    法律状态:
    2018-03-06| B03A| Publication of a patent application or of a certificate of addition of invention [chapter 3.1 patent gazette]|
    2018-10-30| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|
    2020-05-12| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|
    2021-11-09| B03H| Publication of an application: rectification [chapter 3.8 patent gazette]|Free format text: REFERENTA A RPI 2461 DE 06/03/2018,QUANTO AO ITEM (54). |
    优先权:
    申请号 | 申请日 | 专利标题
    IN2155/CHE/2014|2014-04-29|
    IN2155CH2014|2014-04-29|
    [返回顶部]