WO2019165759A1 - 风电场中的风力发电机组的前馈控制方法和设备 - Google Patents

风电场中的风力发电机组的前馈控制方法和设备 Download PDF

Info

Publication number
WO2019165759A1
WO2019165759A1 PCT/CN2018/100245 CN2018100245W WO2019165759A1 WO 2019165759 A1 WO2019165759 A1 WO 2019165759A1 CN 2018100245 W CN2018100245 W CN 2018100245W WO 2019165759 A1 WO2019165759 A1 WO 2019165759A1
Authority
WO
WIPO (PCT)
Prior art keywords
predetermined
feedforward control
prediction
prediction model
wind
Prior art date
Application number
PCT/CN2018/100245
Other languages
English (en)
French (fr)
Inventor
欧发顺
Original Assignee
北京金风科创风电设备有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京金风科创风电设备有限公司 filed Critical 北京金风科创风电设备有限公司
Priority to AU2018411224A priority Critical patent/AU2018411224B2/en
Priority to US16/640,295 priority patent/US11293400B2/en
Priority to EP18907660.7A priority patent/EP3608536B1/en
Priority to ES18907660T priority patent/ES2962474T3/es
Publication of WO2019165759A1 publication Critical patent/WO2019165759A1/zh

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/043—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic
    • F03D7/045—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic with model-based controls
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/0204—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor for orientation in relation to wind direction
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/022—Adjusting aerodynamic properties of the blades
    • F03D7/0224—Adjusting blade pitch
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/028—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor controlling wind motor output power
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/043—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/043—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic
    • F03D7/046—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic with learning or adaptive control, e.g. self-tuning, fuzzy logic or neural network
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/048—Automatic control; Regulation by means of an electrical or electronic controller controlling wind farms
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2260/00—Function
    • F05B2260/82—Forecasts
    • F05B2260/821—Parameter estimation or prediction
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/10—Purpose of the control system
    • F05B2270/103—Purpose of the control system to affect the output of the engine
    • F05B2270/1033—Power (if explicitly mentioned)
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/32—Wind speeds
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/321—Wind directions
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/322—Control parameters, e.g. input parameters the detection or prediction of a wind gust
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/327—Rotor or generator speeds
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/328—Blade pitch angle
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/335—Output power or torque
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/72—Wind turbines with rotation axis in wind direction

Definitions

  • the present application relates generally to the field of wind power generation technology and, more particularly, to a feedforward control method and apparatus for a wind power plant in a wind farm.
  • the fan control system controls the wind turbine to perform the corresponding operation according to the current operating state. Ground action to achieve maximum capture of wind energy by wind turbines.
  • the control command is generated from the fan control system, and when the actuator receives the control command to start the action until the action is completed, it usually needs to last for several control cycles, and in the above control process, the wind speed may be Changes have occurred, and the hysteresis of the fan control system and its actuators is likely to cause over-speed failure of the wind turbine or cause a sharp increase in the load of the wind turbine, which will bring safe operation and long-term fatigue load to the wind turbine. influences.
  • the application provides a feedforward control method and device for a wind power generator in a wind farm, which can realize advanced control of the wind power generator and contribute to safe operation of the wind power generator set.
  • a feedforward control method for a wind power generator in a wind farm including: acquiring real-time operational data of a predetermined wind turbine in a wind farm; The acquired real-time operational data is input to a predetermined prediction model corresponding to the predetermined wind turbine to acquire prediction data through the predetermined prediction model; and determining whether to feed forward the predetermined wind turbine based on the acquired prediction data Controlling a function to control an operating state of the predetermined wind turbine based on a predetermined feedforward control mode.
  • a feedforward control device for a wind power generator in a wind farm
  • the feedforward control device including: a data acquisition module that acquires a predetermined wind turbine in a wind farm Realizing data in real time; a prediction module, inputting the acquired real-time operational data into a predetermined prediction model corresponding to the predetermined wind turbine generator to acquire prediction data through the predetermined prediction model; a feedforward control module, according to the acquired prediction The data determines whether a feedforward control function is turned on for the predetermined wind turbine to control an operational state of the predetermined wind turbine based on a predetermined feedforward control mode.
  • a feedforward control system for a wind power generator in a wind farm including a feedforward control device of a wind power generator in the wind farm described above.
  • a computer readable storage medium storing a computer program that, when executed by a processor, implements feedforward of a wind turbine in the wind farm described above Control Method.
  • a field group controller including: a processor; a memory storing a computer program, when the computer program is executed by the processor, The feedforward control method of the wind turbine in the above wind farm.
  • FIG. 1 shows a flow chart of a feedforward control method of a wind power generator in a wind farm according to an exemplary embodiment of the present application
  • FIG. 2 shows a structural diagram of a feedforward control device of a wind power generator in a wind farm according to an exemplary embodiment of the present application
  • FIG. 3 illustrates a structural diagram of a feedforward control process for a wind speed prediction model according to an exemplary embodiment of the present application
  • FIG. 4 illustrates a structural diagram of a feedforward control process for at least two prediction models, according to an exemplary embodiment of the present application.
  • FIG. 1 illustrates a flow chart of a feedforward control method of a wind power generator in a wind farm according to an exemplary embodiment of the present application.
  • the feedforward control method shown in FIG. 1 can be performed in a field group controller of a wind farm, where the field group controller can be referred to as a wind farm controller (WFC) for controlling the entire wind farm.
  • WFC wind farm controller
  • All wind turbines included in the project can be customized and optimized for wind turbine control to increase the power generation capacity of the wind farm.
  • step S10 real-time operational data of a predetermined wind turbine in a wind farm is acquired.
  • the real-time operational data obtained above may include real-time wind parameters when the wind turbine is scheduled to operate and real-time operating parameters of the predetermined wind turbine itself.
  • the acquired real-time operational data may be a real-time operational data of a predetermined wind turbine based on a time series corresponding to a predetermined prediction model.
  • step S20 the acquired real-time operational data is input to a predetermined prediction model corresponding to the predetermined wind power generator to acquire the predicted data through the predetermined prediction model.
  • the prediction data acquired by the predetermined prediction model may include wind parameters and operational parameters of the predetermined wind turbine itself.
  • the predicted data after the predetermined time may be acquired by the predetermined prediction model, which may be the minimum time required to control the predetermined wind turbine to complete the action corresponding to the feedforward control mode (eg, from generating the control command to completing the control command) A predetermined multiple of the time of the indicated action.
  • the predetermined time may be a second-order time, that is, a predetermined prediction model may be used for short-term prediction.
  • the predicted time length of the predetermined prediction model may be related to the sampling period of the training data for the predetermined prediction model, and the shorter the sampling period of the training data, the shorter the prediction time length of the predetermined prediction model.
  • the predetermined prediction model of the exemplary embodiment of the present application can be used for prediction of a medium-long period (for example, ten seconds, several tens of seconds, several minutes, ten minutes, in addition to being usable for short-time prediction).
  • the real-time running data of the predetermined wind turbine set with a medium-long sampling period (for example, about ten seconds) can be used to obtain the prediction data of the medium-long period (for example, about ten minutes) by using the predetermined prediction model to obtain the medium length.
  • the predicted data of the cycle is supplemented by a wind power prediction system of a longer period (eg, tens of minutes, hours, days).
  • step S30 it is determined whether the feedforward control function is turned on for the predetermined wind turbine based on the acquired predicted data to control the operational state of the predetermined wind turbine based on the predetermined feedforward control mode.
  • Determining whether to turn on the feedforward control function for the predetermined wind turbine based on the acquired prediction data may include determining whether the prediction accuracy of the predetermined prediction model satisfies the requirement. When the prediction accuracy of the predetermined prediction model satisfies the requirement, it is determined that the feedforward control function is turned on for the predetermined wind turbine, and when the prediction accuracy of the predetermined prediction model does not satisfy the requirement, it is determined that the feedforward control function is not turned on for the predetermined wind turbine.
  • the step of determining whether the prediction accuracy of the predetermined prediction model satisfies the requirement may include: inputting the real-time operational data of the predetermined number of sampling periods before the acquired current time to the predetermined prediction model to obtain a plurality of prediction data within the predetermined sampling period, Determining whether the plurality of prediction data and the plurality of actual measurement data corresponding to the plurality of prediction data are consistent, and if the plurality of prediction data are consistent with the plurality of actual measurement data, determining that the prediction of the predetermined prediction model is accurate Degree meets the requirements.
  • real-time running data before the current time may be input to a predetermined prediction model to obtain a reservation.
  • Multiple forecast data after time eg, forecast data from 09:59:57-09:59:59
  • the actual measurement data is data corresponding to the obtained prediction data, that is, if the predicted data is wind speed, the actual measurement data is also the wind speed.
  • the predetermined prediction model may be online trained based on the acquired real-time operational data. And after the training, it is continued to judge whether the prediction accuracy of the predetermined prediction model satisfies the requirement to apply the predetermined prediction model to the feedforward control after the prediction accuracy of the predetermined prediction model satisfies the requirement.
  • Determining whether the plurality of prediction data within the predetermined sampling period and the plurality of actual measurement data respectively corresponding to the plurality of prediction data are consistent may include: calculating a plurality of prediction data within a predetermined sampling period and respectively respectively The average absolute error or the average absolute error percentage of the plurality of actual measurement data corresponding to the prediction data. If the average absolute error is greater than a set threshold or a mean absolute error percentage corresponding to the predetermined prediction model is greater than a set percentage threshold corresponding to the predetermined prediction model, determining that the plurality of prediction data are consistent with the corresponding plurality of actual measurement data That is, it indicates that the prediction accuracy of the predetermined prediction model satisfies the requirement.
  • the data is inconsistent, that is, it indicates that the prediction accuracy of the predetermined prediction model does not meet the requirements.
  • the following formula can be used to calculate the mean absolute error (MAE, Mean Absolute Error),
  • y j is the actual measured data of the jth data sampling point
  • m is the number of data sampling points included in the predetermined sampling period.
  • the following formula can be used to calculate the Mean Absolute Percentage Error (MAPE),
  • the prediction accuracy of the predetermined prediction model may be determined by other means, for example, The prediction accuracy of the predetermined prediction model is determined by calculating SDMAE (standard deviation of mean absolute error), SDMAPE (standard deviation of mean absolute error percentage).
  • the predetermined prediction model may perform online training according to the acquired real-time operational data of the predetermined wind turbine, and after the prediction accuracy of the predetermined prediction model satisfies the requirement, the predetermined prediction model is put into the feedforward.
  • the feedforward control is provided with predictive data. It will be appreciated that existing learning training methods can be utilized to train predetermined predictive models based on acquired real-time operational data of predetermined wind turbines.
  • the predetermined prediction model can also be trained in an offline manner.
  • the predetermined prediction model may be trained based on historical operational data of the predetermined wind turbine, and the trained predetermined prediction model may be tested. When the prediction accuracy of the predetermined prediction model satisfies the requirement, the predetermined prediction model is put into the real-time.
  • the predetermined prediction model is continuously tested online during the feedforward control process to ensure the prediction accuracy of the predicted model in the feedforward control process.
  • the predetermined prediction model may include at least one of the following: a wind speed prediction model, a wind direction prediction model, a turbulence intensity prediction model, and a generator rotation speed prediction model.
  • the predetermined prediction model may be a prediction model, and the prediction data predicted by the one prediction model may be used to determine whether the feedforward control function is turned on for the predetermined wind turbine to be controlled based on a predetermined feedforward control mode. The operational state of the predetermined wind turbine.
  • Determining whether to turn on the feedforward control function for the predetermined wind turbine according to the predicted data may include: determining whether the predicted data change amount within the predetermined time period is greater than a set value, and if greater than, passing the pitch control mode or the electromagnetic torque control mode Control the operational status of the scheduled wind turbine.
  • the real-time operation data corresponding to the wind speed prediction model acquired in step S10 may include: real-time measured wind speed, wind direction, generator speed, output power, and power generation.
  • the electromagnetic torque of the machine, correspondingly predicted by the prediction model, can be the predicted wind speed.
  • a hub front wind measurement system can be installed on the predetermined wind power generator, because the hub front wind measurement system can relatively accurately measure the freedom of the predetermined wind turbine impeller plane at the hub.
  • the wind speed of the flow so using the wind speed measured by the front wind measurement system of the hub to train the wind speed prediction model can greatly improve the accuracy of the wind speed prediction, thereby further expanding the function of the feedforward control, that is, not only for the large turbulence
  • the feedforward control of the wind turbine is scheduled, and can also be used as feedforward control of the predetermined wind turbine under normal operating conditions to turn the predetermined wind turbine into an active control system, thereby greatly improving the operation of the scheduled wind turbine. Efficiency and reduced load under extreme conditions.
  • the wind speed can be calibrated by using the wind speed measured by the wind tower to measure the wind speed measured by the anemometer of the predetermined wind power generator by using the wind tower in the direction of the main wind of the predetermined wind power generator, so that the wind speed is adjusted.
  • the measured wind speed is more accurate to extend the working range of feedforward control and improve the ability of the scheduled wind turbine to actively control.
  • the installation position of the wind tower on the predetermined wind power generator is fixed, when the wind direction changes, the predetermined wind power generator rotates with the wind direction, and the accuracy of the wind speed measured by the wind tower is lowered. That is, the wind speed measured by the anemometer can be calibrated using the wind speed measured by the wind tower in a predetermined range of sectors centered on the main wind direction of the predetermined wind turbine.
  • the following is a step of predicting the wind speed prediction model and predicting the wind speed as an example, and introducing the step of the feedforward control method for the wind speed prediction model.
  • Determining whether to turn on the feedforward control function for the predetermined wind turbine based on the acquired prediction data may include determining whether the amount of change in the predicted wind speed within the predetermined time is greater than a set value (eg, setting the wind speed change amount). If the amount of change in the predicted wind speed (which may be predicted as the amount of change in the predicted increase in wind speed or the amount of change in the predicted wind speed decrease) is greater than the set value, the operation of the predetermined wind turbine may be controlled by a pitch control mode or an electromagnetic torque control mode. status.
  • a set value eg, setting the wind speed change amount
  • the amount of change in the predicted wind speed is greater than the set value, it may be determined based on the predicted wind speed whether the predetermined wind turbine is in a full-shot phase after the predetermined time.
  • whether the predetermined wind turbine is in a full-shot phase may be determined based on a comparison result between the predicted wind speed and the rated wind speed, and when the predicted wind speed is not less than the rated wind speed, the predetermined wind turbine may be considered to be in a full-scale phase, when the predicted wind speed is less than the rated At wind speed, it can be considered that the scheduled wind turbine is in a stage of dissatisfaction.
  • the operational state of the predetermined wind turbine is controlled by the pitch control mode.
  • the pitch control mode As an example, when the predicted wind speed is greater than the rated wind speed, it can be considered that the predetermined wind power generating set is in the full-speed phase at this time, in which case the constant power adjustment can be performed by the pitch control mode, that is, the predetermined wind power generating set is obtained by pitching Constant power output (to keep the wind turbine output at rated power).
  • the predetermined wind power generation set when the amount of change in the predicted wind speed increase is greater than the set wind speed change amount, if the predetermined wind power generation set is in the full-shot phase, the predetermined wind power generation set is controlled to increase the pitch angle, and when the predicted wind speed decrease is greater than the set When the wind speed varies, if the predetermined wind turbine is in the full-shot phase, the predetermined wind turbine is controlled to reduce the pitch angle.
  • the operational state of the predetermined wind turbine can be controlled by electromagnetic torque control.
  • electromagnetic torque control such as increasing the electromagnetic torque
  • the generator speed is adjusted, the blade is operated at the optimal tip speed ratio, and the wind energy utilization coefficient of the blade is the largest (Cpmax) to achieve the predetermined wind power. The largest capture of wind energy by the generator set.
  • the predetermined wind power generator set is in the dissatisfaction phase, the predetermined wind power generator is controlled to increase the electromagnetic torque, and when the predicted wind speed decrease is greater than the set wind speed change amount At the time, if the predetermined wind turbine is in a dissatisfaction phase, the predetermined wind turbine is controlled to reduce the electromagnetic torque.
  • the real-time operation data corresponding to the wind direction prediction model acquired in step S10 may include: real-time measured wind speed, wind direction, cabin position, generator speed, and output. Power, electromagnetic torque of the generator.
  • the real-time operation data corresponding to the turbulence intensity prediction model acquired in step S10 may include: real-time measured wind speed, wind direction, and generator speed. , output power, electromagnetic torque of the generator.
  • the turbulence intensity is a ratio of the standard deviation of the wind speed of the predetermined time period to the average wind speed of the predetermined time period, that is, the turbulence intensity is an estimated value of the predetermined time period, and the turbulence predicted by the turbulence intensity prediction model
  • the intensity may not be used as feedforward control, but only the predicted turbulence intensity is provided to the operator of the wind farm for the operator to understand the trend of turbulence intensity.
  • the real-time operation data corresponding to the generator rotation speed prediction model acquired in step S10 may include: real-time measured wind speed, wind direction, generator speed, The output power, the electromagnetic torque of the generator, the acceleration of the wind turbine in the first predetermined direction (X direction) and the second predetermined direction (Y direction).
  • the first predetermined direction may refer to the direction from the head to the tail of the wind turbine
  • the second predetermined direction refers to the direction perpendicular to the wind direction (eg, the field worker stands in the downwind direction, facing the nose, the field staff
  • the left and right direction can be defined as a second predetermined direction).
  • the generator rotational speed prediction model can be used to predict the generator rotational speed to adjust in advance based on the predicted generator rotational speed through the pitch control mode to avoid generator overspeed faults and reduce power generation. Loss.
  • the predetermined prediction model may include at least two prediction models.
  • the feedforward control method of the wind turbine in the wind farm may further include determining whether the comprehensive accuracy of the at least two prediction models satisfies the requirements.
  • the step of determining whether the comprehensive accuracy of the at least two prediction models satisfies the requirements may include: setting a weight value for each of the at least two prediction models, based on the set weight value and the prediction accuracy of each prediction model The comprehensive accuracy is determined. When the comprehensive accuracy is greater than the preset threshold, the comprehensive accuracy is determined to meet the requirement. When the comprehensive accuracy is not greater than the preset threshold, it is determined that the comprehensive accuracy does not meet the requirement.
  • the following formula can be used to calculate the combined accuracy of at least two predictive models
  • the step of determining whether the comprehensive accuracy of the at least two prediction models satisfies the requirements may include: determining prediction accuracy of each prediction model, and determining comprehensive accuracy when the prediction accuracy of each prediction model satisfies the requirements Degree meets the requirements.
  • the yaw control needs to determine the yaw angle according to the wind speed and the wind direction to control the yaw angle determined by the rotation of the wind turbine to achieve wind operation.
  • the predetermined prediction model includes a wind speed predetermined model and a wind direction prediction model
  • the yaw angle is determined according to the wind speed predicted by the wind speed predetermined model and the wind direction predicted by the wind direction prediction model to control the rotation of the predetermined wind turbine set by the yaw control method. Yaw angle.
  • FIG. 2 shows a structural diagram of a feedforward control device of a wind power generator in a wind farm according to an exemplary embodiment of the present application.
  • the feedforward control apparatus 100 of a wind power generator in a wind farm includes a data acquisition module 10, a prediction module 20, and a feedforward control module 30.
  • the data acquisition module 10 acquires real-time operational data of a predetermined wind turbine in the wind farm.
  • the acquired real-time operational data may include real-time wind parameters for the scheduled wind turbine operation and real-time operational parameters of the predetermined wind turbine itself.
  • the acquired real-time operational data may be a real-time operational data of a predetermined wind turbine based on a time series corresponding to a predetermined prediction model.
  • the prediction module 20 inputs the acquired real-time operational data to a predetermined prediction model corresponding to the predetermined wind turbine to acquire the predicted data through the predetermined prediction model.
  • the prediction data acquired by the predetermined prediction model may include wind parameters and operational parameters of the predetermined wind turbine itself.
  • the prediction module 20 may also acquire prediction data after a predetermined time by a predetermined prediction model, which may be a predetermined multiple of a minimum time required to control a predetermined wind turbine to complete an action corresponding to the feedforward control mode.
  • a predetermined prediction model which may be a predetermined multiple of a minimum time required to control a predetermined wind turbine to complete an action corresponding to the feedforward control mode.
  • the feedforward control module 30 determines whether to turn on the feedforward control function for the predetermined wind turbine based on the acquired prediction data to control the operational state of the predetermined wind turbine based on a predetermined feedforward control mode.
  • the feedforward control device of the wind turbine in the wind farm may further include: a test module (not shown) that determines whether the predicted accuracy of the predetermined prediction model satisfies the requirement.
  • the test module inputs real-time operational data of the acquired predetermined sampling period (which may be within a predetermined sampling period before the current time) to a predetermined prediction model to obtain a plurality of prediction data within the predetermined sampling period, and determines the location. Determining whether the plurality of prediction data and the plurality of actual measurement data corresponding to the plurality of prediction data are consistent, and if the plurality of prediction data are consistent with the plurality of actual measurement data, determining that the prediction accuracy of the predetermined prediction model is satisfied Claim.
  • the feedforward control module 30 determines to turn on the feedforward control function for the predetermined wind turbine, and if the prediction accuracy of the predetermined prediction model does not satisfy the requirement, the feedforward control module 30 determines that the predetermined The wind turbine does not turn on feedforward control.
  • the test module may perform online training on the predetermined prediction model based on the acquired real-time operational data. And after the training, it is continued to judge whether the prediction accuracy of the predetermined prediction model satisfies the requirement to apply the predetermined prediction model to the feedforward control after the prediction accuracy of the predetermined prediction model satisfies the requirement.
  • the test module may calculate a plurality of prediction data within a predetermined sampling period and an average absolute error or a mean absolute error percentage of the plurality of actual measurement data respectively corresponding to the plurality of prediction data. If the average absolute error is greater than a set threshold or a mean absolute error percentage corresponding to the predetermined prediction model is greater than a set percentage threshold corresponding to the predetermined prediction model, the test module determines the plurality of prediction data and the corresponding plurality of actual measurements The data is consistent, that is, indicating that the prediction accuracy of the predetermined prediction model satisfies the requirements.
  • the data is inconsistent, that is, it indicates that the prediction accuracy of the predetermined prediction model does not meet the requirements.
  • the predetermined prediction model may perform online training based on the acquired real-time operational data of the predetermined wind turbine, or may also perform offline training on the predetermined prediction model based on historical operational data of the predetermined wind turbine.
  • the predetermined prediction model may include at least one of the following: a wind speed prediction model, a wind direction prediction model, a turbulence intensity prediction model, and a generator rotation speed prediction model.
  • the predetermined prediction model may be a prediction model.
  • the feedforward control module 30 may determine whether to turn on the feedforward control function for the predetermined wind turbine according to the prediction data predicted by the one prediction model, so as to be based on the predetermined The feedforward control mode controls the operating state of the predetermined wind turbine.
  • the feedforward control module 30 determines whether the amount of change in the predicted data within the predetermined time period is greater than a set value, and if the amount of change in the predicted data is greater than the set value, the feedforward control module 30 determines to turn on the feedforward control function for the predetermined wind turbine set so that The operating state of the predetermined wind turbine is controlled by a pitch control mode or an electromagnetic torque control mode.
  • the exemplary embodiment of the present application further provides a feedforward control system for a wind power generator in a wind farm, which may include a plurality of units in addition to the feedforward control device 100 shown in FIG. Controller 200.
  • Each unit controller 200 can control the operational status of the respective wind turbine.
  • the unit controller 200 corresponding to the predetermined wind power generator group can control the predetermined wind power generation by the pitch control mode or the electromagnetic torque control mode. The operating status of the unit.
  • the real-time operation data corresponding to the wind speed prediction model acquired by the data acquisition module 10 may include: real-time measured wind speed, wind direction, generator speed, and output.
  • the power, the electromagnetic torque of the generator, and the predicted data predicted by the corresponding prediction model may be the predicted wind speed.
  • the following is a process in which the predetermined prediction model is the wind speed prediction model and the prediction data is the predicted wind speed. Referring to FIG. 2, the process of the feedforward control mode for the wind speed prediction model is introduced.
  • FIG. 3 illustrates a structural diagram of a feedforward control process for a wind speed prediction model according to an exemplary embodiment of the present application.
  • the feedforward control device 100 of FIG. 2 shown in the above exemplary embodiment of the present application may include a data acquisition module corresponding to each wind power generator set, a prediction module and a test module respectively corresponding to each prediction model, And a feedforward control module.
  • the feedforward control process for the wind speed prediction model may be: the data acquisition module 1 acquires the real-time wind speed of the wind turbine 1 during operation, and the prediction module 1 runs the acquired wind turbine 1 The real-time wind speed at the time is input to the wind speed prediction model 1 corresponding to the wind turbine 1, and the predicted wind speed of the wind turbine 1 is obtained.
  • the test module 1 determines whether the prediction accuracy of the wind speed prediction model 1 satisfies the requirement.
  • the test module 1 sends an enable signal to the feed forward control module, and the feedforward control module responds to the The enable signal determines that the feedforward control function is turned on for the wind turbine 1 and issues a control command to the unit controller 1 according to the predicted wind speed predicted by the wind speed prediction model 1, and the unit controller 1 adjusts the electromagnetic torque in response to the control command or Pitch in advance.
  • test module 1 determines that the prediction accuracy of the wind speed prediction model 1 does not meet the requirement, the test module 1 does not send an enable signal to the feed forward control module, and the feedforward control module determines that the feedforward control function is not turned on for the wind turbine 1 , indicating that the wind speed prediction model 1 at this time needs further learning to adapt to the current changing wind conditions.
  • the real-time running data of the wind turbine 1 can be obtained in real time, and the wind speed prediction module 1 corresponding to the wind turbine 1 is trained in real time, and the test module 1 performs real-time testing.
  • the feedforward control module is enabled.
  • the feedforward control module can push different control parameters to each unit controller to avoid the occurrence of each wind turbine. Over-speed failure or by pitching in advance to reduce the load on the wind turbine makes the wind turbine more efficient and less loaded.
  • the feedforward control module 30 may determine whether the amount of change in the predicted wind speed within a predetermined time is greater than a set value (eg, set a wind speed change amount). If the amount of change in the predicted wind speed (which may be predicted as the amount of change in the predicted increase in wind speed or the amount of change in the predicted wind speed decrease) is greater than the set value, the unit controller corresponding to the predetermined wind turbine may be controlled by pitch control or electromagnetic torque. The mode controls the operating state of the predetermined wind turbine.
  • a set value eg, set a wind speed change amount
  • the feedforward control module 30 may determine whether the predetermined wind turbine is in a full-shot phase after the predetermined time based on the predicted wind speed.
  • the feedforward control module 30 may determine whether the predetermined wind turbine is in a full-shot phase based on a comparison result between the predicted wind speed and the rated wind speed, and when the predicted wind speed is not less than the rated wind speed, the feedforward control module 30 may consider that the predetermined wind turbine is at In the full-scale phase, when the predicted wind speed is less than the rated wind speed, the feedforward control module 30 may consider that the predetermined wind turbine is in a stage of dissatisfaction.
  • the feedforward control module 30 may determine to turn on the feedforward control function for the predetermined wind turbine, and determine that the feedforward control mode at this time is the pitch control mode, corresponding to the predetermined wind turbine
  • the unit controller can control the running state of the predetermined wind turbine by means of pitch control.
  • the unit controller corresponding to the predetermined wind power generator group can perform constant power adjustment through the pitch control mode, that is, The pitch makes the predetermined wind turbine constant power output (to keep the wind turbine output at rated power).
  • the unit controller corresponding to the predetermined wind power generating set may control the predetermined wind power generating set to increase the pitch angle, when predicting When the amount of change in the wind speed decrease is greater than the set wind speed change amount, if the predetermined wind power generator set is in the full-shot phase, the unit controller corresponding to the predetermined wind power generator set can control the predetermined wind power generator set to reduce the pitch angle.
  • the feedforward control module 30 may determine to turn on the feedforward control function for the predetermined wind turbine, and determine that the feedforward control mode at this time is an electromagnetic torque control mode, corresponding to the predetermined wind turbine
  • the unit controller can control the running state of the predetermined wind turbine by electromagnetic torque control.
  • the predetermined wind power generating set is in the stage of dissatisfaction, that is, in the maximum wind energy capturing stage, in which case the blade pitch angle is maintained at the optimal pitch angle, and
  • the unit controller corresponding to the scheduled wind turbine generator can adjust the electromagnetic torque of the generator by electromagnetic torque control (such as increasing the electromagnetic torque), adjust the generator speed, and make the blade run at the optimal tip speed ratio, and the wind energy utilization coefficient of the blade. Maximum to achieve maximum capture of wind energy by a predetermined wind turbine.
  • the unit controller corresponding to the predetermined wind power generating set may control the predetermined wind power generating set to increase the electromagnetic torque.
  • the unit controller corresponding to the predetermined wind power generating set may control the predetermined wind power generating set to reduce the electromagnetic torque.
  • the real-time operational data corresponding to the wind direction prediction model acquired by the data acquisition module 10 may include: real-time measured wind speed, wind direction, cabin position, generator speed, and output. Power, electromagnetic torque of the generator.
  • the real-time operational data corresponding to the turbulence intensity prediction model acquired by the data acquisition module 10 may include: real-time measured wind speed, wind direction, and generator speed. , output power, electromagnetic torque of the generator.
  • the turbulence intensity is a ratio of the standard deviation of the wind speed of the predetermined time period to the average wind speed of the predetermined time period, that is, the turbulence intensity is an estimated value of the predetermined time period, and the turbulence predicted by the turbulence intensity prediction model
  • the intensity may not be used as feedforward control, but only the predicted turbulence intensity is provided to the operator of the wind farm for the operator to understand the trend of turbulence intensity.
  • the feedforward control system of the wind power generator in the wind farm may further include: a display for displaying the predicted data obtained using the predetermined prediction model.
  • the feedforward control module can transmit the turbulence intensity predicted by the turbulence intensity prediction model to the display for display.
  • the predetermined prediction model is a generator rotation speed prediction model.
  • the real-time operation data corresponding to the generator rotation speed prediction model acquired by the data acquisition module 10 may include: real-time measured wind speed, wind direction, generator speed, The output power, the electromagnetic torque of the generator, the acceleration of the wind turbine in the first predetermined direction (X direction) and the second predetermined direction (Y direction).
  • the generator rotational speed can be predicted by the generator rotational speed prediction model, and the unit controller corresponding to the wind turbine generating set corresponding to the generator rotational speed prediction model passes the pitch control based on the predicted generator rotational speed.
  • the method is to adjust in advance to avoid the occurrence of generator overspeed faults and reduce the loss of power generation.
  • the predetermined prediction model may include at least two prediction models.
  • the process of the feedforward control mode for at least two prediction models is described below with reference to FIG.
  • FIG. 4 illustrates a structural diagram of a feedforward control process for at least two prediction models, according to an exemplary embodiment of the present application.
  • the feedforward control apparatus 100 in FIG. 2 may include a data acquisition module corresponding to each wind turbine, a prediction module corresponding to each prediction model, a test module and a scheduling module, and a feedforward control module.
  • test module can be used to determine whether the prediction accuracy of the prediction model corresponding to each wind turbine meets the requirements.
  • the scheduling module can be used to determine whether the combined accuracy of the at least two predictive models meets the requirements.
  • the scheduling module may set a weight value for each of the at least two prediction models, and determine a comprehensive accuracy based on the set weight value and the prediction accuracy of each prediction model, when the comprehensive accuracy is greater than When the threshold is preset, the comprehensive accuracy is determined to meet the requirement. When the comprehensive accuracy is not greater than the preset threshold, it is determined that the comprehensive accuracy does not meet the requirement.
  • the scheduling module determines that the overall accuracy meets the requirements when the prediction accuracy of each of the prediction models satisfies the requirements.
  • the scheduling module determines that the comprehensive accuracy of the at least two prediction models meets the requirements, the scheduling module sends an enable signal to the feedforward control module, and the feedforward control module determines, in response to the enable signal, that the predetermined wind turbine is turned on. Feed control function.
  • the scheduling module determines that the comprehensive accuracy of the at least two prediction models does not meet the requirement, the scheduling module does not send an enable signal to the forward feed control module, and the feedforward control module determines that the feedforward control is not turned on for the predetermined wind turbine Function, indicating that the at least two predictive models at this time require further learning.
  • the yaw control needs to determine the yaw angle according to the wind speed and the wind direction to control the yaw angle determined by the rotation of the predetermined wind turbine to achieve the wind operation.
  • the predetermined prediction model includes the wind speed predetermined model and the wind direction prediction model
  • the feedforward control module predicts the wind direction according to the wind speed and wind direction prediction model predicted by the wind speed predetermined model.
  • the unit controller corresponding to the predetermined wind turbine controls the yaw angle determined by the rotation of the predetermined wind turbine by the yaw control mode.
  • a computer readable storage medium storing a computer program is also provided in accordance with an exemplary embodiment of the present application.
  • the computer readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform a feedforward control method of the wind turbine in the wind farm described above.
  • the computer readable recording medium is any data storage device that can store data read by a computer system. Examples of the computer readable recording medium include read only memory, random access memory, read-only optical disk, magnetic tape, floppy disk, optical data storage device, and carrier wave (such as data transmission via the Internet via a wired or wireless transmission path).
  • a field group controller is also provided in accordance with an exemplary embodiment of the present application.
  • the field group controller includes a processor and a memory.
  • the memory is used to store computer programs.
  • the computer program is executed by a processor such that the processor performs a feedforward control method of the wind turbine in the wind farm as described above.
  • the feedforward control method and apparatus for a wind power generator in a wind farm can prevent the wind turbine from being subjected to extreme conditions by predicting the operation state of the wind turbine at a future time to control the wind turbine. The impact of operation and load.
  • the feedforward control method and apparatus of the wind power generator in the wind farm of the exemplary embodiment of the present application are combined with the feedforward control method to change the unit control system from the passive control system to some extent. It has become an active control system, and enables the unit control system to sense changes in wind speed, wind direction, generator speed, and turbulence intensity in advance, and generate corresponding control commands in advance, eliminating the lag of control brought by the current passive system of the unit.
  • the wind turbine output can be improved to avoid over-speed or acceleration over-limit failure of the wind turbine.
  • the feedforward control method and apparatus for a wind power generator in a wind farm by predicting a large turbulence condition, issue a control command to the wind power generator in advance (such as increasing electromagnetic torque) to make electromagnetic torque Balance with pneumatic torque to avoid sudden increase in wind turbine speed and improve the output of wind turbines while avoiding over-speed faults, reducing the load on wind turbines.
  • a feedforward control method and apparatus for a wind power generator in a wind farm is used to predict a sudden increase in wind speed by using a wind speed prediction module when the wind power generation unit is in a full-scale phase, Wind power generator constant power control is achieved by pre-pitching to unload the wind speed sudden increase of the load applied to the wind turbine.
  • the current forecasting speed of the wind speed power prediction system adopted in China is 10-20 min, that is, the wind speed and power value of the wind farm are predicted every 10-20 min to guide the grid dispatching.
  • the above-mentioned predetermined prediction model of the exemplary embodiment of the present application may be used to perform long-term prediction, such as predicting wind speed and power values in the range of 5-20 min, to predict the wind speed and power values as current wind speed power predictions. An important addition to the system.

Abstract

一种风电场中的风力发电机组的前馈控制方法，包括：获取风电场中的预定风力发电机组的实时运行数据；将获取的实时运行数据输入到与预定风力发电机组对应的预定预测模型，以通过预定预测模型获取预测数据；根据获取的预测数据确定是否针对预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制预定风力发电机组的运行状态。该方法能够实现对风力发电机组的提前控制，有助于风力发电机组的安全运行。还涉及一种风电场中的风力发电机组的前馈控制设备，控制系统，计算机可读存储介质以及场群控制器。

Description

风电场中的风力发电机组的前馈控制方法和设备 技术领域
本申请总体说来涉及风力发电技术领域，更具体地讲，涉及一种风电场中的风力发电机组的前馈控制方法和设备。
背景技术
风力发电机组在正常运行过程中，由于外界风速的变化，风力发电机组的运行状态随之发生变化，导致叶片吸收的气动功率发生变化，风机控制系统根据当前的运行状态，控制风力发电机组执行相应地动作，以实现风力发电机组对风能的最大捕获。在此过程中，由于风速变化较快，从风机控制系统产生控制指令，到执行机构接收到控制指令开始动作，直至完成动作，通常需要持续数个控制周期，而在上述控制过程中，风速可能已经发生了变化，风机控制系统及其执行机构的滞后性很可能导致风力发电机组发生过速故障，或者导致风力发电机组的载荷急剧升高，对风力发电机组安全运行和长期的疲劳载荷带来影响。
发明内容
本申请提供一种风电场中的风力发电机组的前馈控制方法和设备，能够实现对风力发电机组的提前控制，有助于风力发电机组的安全运行。
根据本申请示例性实施例的一方面，提供一种风电场中的风力发电机组的前馈控制方法，所述前馈控制方法包括：获取风电场中的预定风力发电机组的实时运行数据；将获取的实时运行数据输入到与所述预定风力发电机组对应的预定预测模型，以通过所述预定预测模型获取预测数据；根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制所述预定风力发电机组的运行状态。
根据本申请示例性实施例的另一方面，提供一种风电场中的风力发电机组的前馈控制设备，所述前馈控制设备包括：数据获取模块，获取风电场中的预定风力发电机组的实时运行数据；预测模块，将获取的实时运行数据输 入到与所述预定风力发电机组对应的预定预测模型，以通过所述预定预测模型获取预测数据；前馈控制模块，根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制所述预定风力发电机组的运行状态。
根据本申请示例性实施例的再一方面，提供一种风电场中的风力发电机组的前馈控制系统，所述前馈控制系统包括上述的风电场中的风力发电机组的前馈控制设备。
根据本申请示例性实施例的再一方面，提供一种存储有计算机程序的计算机可读存储介质，当所述计算机程序在被处理器执行时实现上述的风电场中的风力发电机组的前馈控制方法。
根据本申请示例性实施例的再一方面，提供一种场群控制器，所述场群控制器包括：处理器；存储器，存储有计算机程序，当所述计算机程序被处理器执行时，实现上述的风电场中的风力发电机组的前馈控制方法。
附图说明
图1示出根据本申请示例性实施例的风电场中的风力发电机组的前馈控制方法的流程图；
图2示出根据本申请示例性实施例的风电场中的风力发电机组的前馈控制设备的结构图；
图3示出根据本申请示例性实施例的针对风速预测模型的前馈控制过程的结构示意图；
图4示出根据本申请示例性实施例的针对至少两个预测模型的前馈控制过程的结构示意图。
具体实施方式
现在，将参照附图更充分地描述不同的示例实施例，一些示例性实施例在附图中示出。
图1示出根据本申请示例性实施例的风电场中的风力发电机组的前馈控制方法的流程图。优选地，图1所示的前馈控制方法可在风电场的场群控制器中执行，这里，场群控制器可指风电场控制器(WFC，wind farm controller)，用于控制整个风电场中包括的所有风力发电机组，可实现风力发电机组控制 的定制化和最优化，以提高风电场的发电能力。
参照图1，在步骤S10中，获取风电场中的预定风力发电机组的实时运行数据。
上述获取的实时运行数据可包括预定风力发电机组运行时的实时风参数和该预定风力发电机组自身的实时运行参数。作为示例，获取的实时运行数据可为预定风力发电机组基于时间序列的与预定预测模型相应的实时运行数据。
在步骤S20中，将获取的实时运行数据输入到与预定风力发电机组对应的预定预测模型，以通过所述预定预测模型获取预测数据。优选地，通过预定预测模型所获取的预测数据可包括风参数和预定风力发电机组自身的运行参数。
通过预定预测模型可获取预定时间之后的预测数据，所述预定时间可为控制预定风力发电机组完成与前馈控制方式对应的动作所需的最短时间(例如，从产生控制指令到完成该控制指令指示的动作的时间)的预定倍数。作为示例，该预定时间可为秒级的时间，也就是说，预定预测模型可用于短时间的预测。应理解，预定预测模型的预测时间长度可与用于该预定预测模型的训练数据的采样周期有关，训练数据的采样周期越短，则该预定预测模型的预测时间长度越短。
本申请示例性实施例的预定预测模型除可用于短时间的预测之外，还可用于中长周期(例如，十几秒、几十秒、几分钟、十几分钟)的预测。可利用中长采样周期(例如，十秒左右)的预定风力发电机组的实时运行数据，通过上述预定预测模型来获得中长周期(例如，十分钟左右)的预测数据，以将获得的中长周期的预测数据作为更长周期(例如，几十分钟、几小时、几天)的风功率预测系统的补充。
在步骤S30中，根据获取的预测数据确定是否针对预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制预定风力发电机组的运行状态。
根据获取的预测数据确定是否针对预定风力发电机组开启前馈控制功能的步骤可包括：判断预定预测模型的预测准确度是否满足要求。当预定预测模型的预测准确度满足要求时，确定针对预定风力发电机组开启前馈控制功能，当预定预测模型的预测准确度不满足要求时，确定针对预定风力发电机 组不开启前馈控制功能。
判断预定预测模型的预测准确度是否满足要求的步骤可包括：将获取的当前时刻之前的预定数量的采样周期的实时运行数据输入到预定预测模型，以获得预定采样周期内的多个预测数据，判断所述多个预测数据和与所述多个预测数据对应的多个实际测量数据是否一致，如果所述多个预测数据与所述多个实际测量数据一致，则确定预定预测模型的预测准确度满足要求。
作为示例，可假设当前时刻为10:00:00，可将当前时刻之前的实时运行数据(例如，09:59:54-09:59:56的实时运行数据)输入到预定预测模型以获得预定时间之后的多个预测数据(如，09:59:57-09:59:59的预测数据)，将09:59:57-09:59:59的多个预测数据与09:59:57-09:59:59的与所述多个预测数据对应的多个实际测量数据进行比较，当所述多个预测数据与所述多个实际测量数据一致时，确定预定预测模型的预测准确度满足要求。应理解，实际测量数据是与所获得的预测数据对应的数据，即，如果预测数据为风速，则实际测量数据也为风速。
此外，如果所述多个预测数据与所述多个实际测量数据不一致，则表明预定预测模型的预测准确度不满足要求，此时可基于获取的实时运行数据对该预定预测模型进行在线训练，并在训练之后继续判断预定预测模型的预测准确度是否满足要求，以在预定预测模型的预测准确度满足要求之后将预定预测模型应用到前馈控制中。
判断预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据是否一致的步骤可包括：计算预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据的平均绝对误差或平均绝对误差百分比。如果平均绝对误差大于与预定预测模型对应的设定阈值或平均绝对误差百分比大于与预定预测模型对应的设定百分比阈值，则确定所述多个预测数据与对应的所述多个实际测量数据一致，即，表明预定预测模型的预测准确度满足要求。如果平均绝对误差不大于与预定预测模型对应的设定阈值或平均绝对误差百分比不大于与预定预测模型对应的设定百分比阈值，则确定所述多个预测数据与对应的所述多个实际测量数据不一致，即，表明预定预测模型的预测准确度不满足要求。
在一些实施例中，可利用下面的公式来计算平均绝对误差(MAE，Mean Absolute Error)，
Figure PCTCN2018100245-appb-000001
公式(1)中，y j为第j个数据采样点的实际测量数据，
Figure PCTCN2018100245-appb-000002
为第j个数据采样点的预测数据，1≤j≤m，m为预定采样周期内包含的数据采样点的个数。
在一些实施例中，可利用下面的公式来计算平均绝对误差百分比(MAPE，Mean Absolute Percentage Error)，
Figure PCTCN2018100245-appb-000003
应理解，除上述通过计算平均绝对误差MAE和平均绝对误差百分比MAPE来确定预定预测模型的预测准确度之外，还可通过其他方式来确定预定预测模型的预测准确度是否满足要求，例如，可通过计算SDMAE(平均绝对误差的标准差)、SDMAPE(平均绝对误差百分比的标准差)来确定预定预测模型的预测准确度。
在一些实施例中，上述预定预测模型可依据获取的预定风力发电机组的实时运行数据来进行在线训练，并当该预定预测模型的预测准确度满足要求之后，将该预定预测模型投入到前馈控制中，以为前馈控制提供预测数据。应理解，可利用现有的各种学习训练方法来基于获取的预定风力发电机组的实时运行数据来对预定预测模型进行训练。
应理解，除上述在线训练预定预测模型的方式之外，还可通过离线方式训练预定预测模型。例如，可基于预定风力发电机组的历史运行数据对预定预测模型进行训练，并对训练后的预定预测模型进行测试，当预定预测模型的预测准确度满足要求时，将该预定预测模型投入上述实时在线的前馈控制中，并在前馈控制过程中对预定预测模型不断进行在线测试，以确保在前馈控制过程中预定预测模型的预测精度。
作为示例，预定预测模型可包括以下至少一项：风速预测模型、风向预测模型、湍流强度预测模型、发电机转速预测模型。
第一种情况，预定预测模型可为一个预测模型，此时可根据该一个预测模型所预测的预测数据来确定是否针对预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制该预定风力发电机组的运行状态。
根据预测数据确定是否针对预定风力发电机组开启前馈控制功能的步骤可包括：判断预定时间段内的预测数据变化量是否大于设定值，如果大于，则通过变桨控制方式或电磁扭矩控制方式控制预定风力发电机组的运行状态。
在第一示例中，以预定预测模型为风速预测模型为例，在步骤S10中获取的与风速预测模型相应的实时运行数据可包括：实时测量的风速、风向、发电机转速、输出功率、发电机的电磁扭矩，相应地预测模型所预测的预测数据可为预测风速。
为提高对风速测量的准确性，可在预定风力发电机组上安装轮毂前置测风系统，由于该轮毂前置测风系统能够比较准确地测量出预定风力发电机组叶轮平面位于轮毂处的自由来流的风速，因此利用该轮毂前置测风系统所测得的风速来训练风速预测模型，可大大提高风速预测的准确性，从而可以进一步扩展前馈控制的功能，即不仅仅用于大湍流情况时预定风力发电机组的前馈控制，还可以作为正常运行条件下的预定风力发电机组的前馈控制，以将预定风力发电机组变为一个主动控制系统，从而大大提高预定风力发电机组的运行效率和降低在极端工况下的载荷。
除此之外，还可通过在预定风力发电机组的主风方向上安装测风塔，以利用测风塔测得的风速对预定风力发电机组的风速仪所测得的风速进行标定，使得风速仪测得的风速更为准确，以扩展前馈控制的工作范围，提高预定风力发电机组主动控制的能力。这里，由于测风塔在预定风力发电机组上的安装位置是固定的，当风向改变时，预定风力发电机组随风向进行转动，此时测风塔所测得的风速的准确性会降低，也就是说，在以预定风力发电机组的主风向为中心的预定范围的扇区内，可利用测风塔所测得的风速对风速仪所测得的风速进行标定。
下面以预定预测模型为风速预测模型，预测数据为预测风速为例，介绍针对风速预测模型的前馈控制方法的步骤。
根据获取的预测数据确定是否针对预定风力发电机组开启前馈控制功能的步骤可包括：确定预定时间内的预测风速的变化量是否大于设定值(例如，设定风速变化量)。如果预测风速的变化量(可指预测风速增加的变化量或预测风速减小的变化量)大于设定值，则可通过变桨控制方式或电磁扭矩控制方式控制所述预定风力发电机组的运行状态。
如果预测风速的变化量大于设定值，则可基于预测风速确定预定风力发电机组在所述预定时间之后是否处于满发阶段。作为示例，可基于预测风速与额定风速的比较结果来确定预定风力发电机组是否处于满发阶段，当预测风速不小于额定风速时，可认为预定风力发电机组处于满发阶段，当预测风 速小于额定风速时，可认为预定风力发电机组处于不满发阶段。
如果预定风力发电机组处于满发阶段，则通过变桨控制方式控制预定风力发电机组的运行状态。作为示例，当预测风速大于额定风速时，可认为此时预定风力发电机组处于满发阶段，在此情况下，可通过变桨控制方式进行恒功率调整，即，通过变桨使预定风力发电机组恒功率输出(使风力发电机组保持以额定功率输出)。例如，当预测风速增加的变化量大于设定风速变化量时，如果预定风力发电机组处于满发阶段，则控制预定风力发电机组增大桨距角，当预测风速减小的变化量大于设定风速变化量时，如果预定风力发电机组处于满发阶段，则控制预定风力发电机组减小桨距角。
如果预定风力发电机组处于不满发阶段，则可通过电磁扭矩控制方式控制预定风力发电机组的运行状态。作为示例，当预测风速不大于额定风速时，可认为此时预定风力发电机组处于不满发阶段，即，处于最大风能捕获阶段，在此情况下，叶片桨距角保持在最优桨距角，可通过电磁扭矩控制方式调整发电机电磁扭矩(如增大电磁扭矩)，调整发电机转速，使叶片运行在最优叶尖速比处，叶片的风能利用系数最大(Cpmax)，以实现预定风力发电机组对风能的最大捕获。当预测风速增加的变化量大于设定风速变化量时，如果预定风力发电机组处于不满发阶段，则控制预定风力发电机组增大电磁扭矩，当预测风速减小的变化量大于设定风速变化量时，如果预定风力发电机组处于不满发阶段，则控制预定风力发电机组减小电磁扭矩。
在第二示例中，以预定预测模型为风向预测模型为例，在步骤S10中获取的与风向预测模型相应的实时运行数据可包括：实时测量的风速、风向、机舱位置、发电机转速、输出功率、发电机的电磁扭矩。
在第三示例中，以预定预测模型为湍流强度预测模型为例，此时，在步骤S10中获取的与湍流强度预测模型相应的实时运行数据可包括：实时测量的风速、风向、发电机转速、输出功率、发电机的电磁扭矩。
在本申请实施例中，湍流强度为预定时间段的风速标准差与该预定时间段的平均风速的比值，也就是说，湍流强度为预定时间段的估计值，湍流强度预测模型所预测的湍流强度可不作为前馈控制使用，而仅将预测的湍流强度提供给风电场的操作人员，以供操作人员了解湍流强度的变化趋势。
在第四示例中，以预定预测模型为发电机转速预测模型为例，在步骤S10中获取的与发电机转速预测模型相应的实时运行数据可包括：实时测量的风 速、风向、发电机转速、输出功率、发电机的电磁扭矩、风力发电机组在第一预定方向(X方向)和第二预定方向(Y方向)上的加速度。作为示例，第一预定方向可指从风力发电机组的头部至尾部的方向，第二预定方向是指与风向垂直的方向(例如，现场工作人员站在下风向，面向机头，现场工作人员的左右方向可定义为第二预定方向)。
由于当风电场中的风力发电机组发生发电机过速故障时，一般无法通过远程控制的方式对故障进行复位，需要风电场的操作人员在风力发电机组的运行现场进行操作才能复位故障。这会使得恢复上述发电机过速故障所需的故障处理时间较长，造成发电量的损失。在本申请一些实施例中可通过发电机转速预测模型来预测发电机转速，以基于预测的发电机转速通过变桨控制方式来提前进行调整，以避免发电机过速故障的发生，减少发电量的损失。
第二种情况，预定预测模型可包括至少两个预测模型。在此情况下，风电场中的风力发电机组的前馈控制方法还可包括：确定所述至少两个预测模型的综合准确度是否满足要求。
确定至少两个预测模型的综合准确度是否满足要求的步骤可包括：为所述至少两个预测模型中的每个预测模型设置权重值，基于设置的权重值和每个预测模型的预测准确度确定综合准确度，当综合准确度大于预设阈值时，确定综合准确度满足要求，当综合准确度不大于预设阈值时，确定综合准确度不满足要求。
可利用下面的公式来计算至少两个预测模型的综合准确度，
Pall＝w 1·p(f(x 1))+w 2·p(f(x 2))+…+w n·p(f(x n))  (3)
公式(3)中，w i为与第i个预测模型对应的权重，1≤i≤n，n为风电场中包括的风力发电机组的数量，且满足：w 1+w 2+…+w n＝1，p(f(x i))为第i个预测模型的预测准确度，x i表示与第i个预测模型对应的实时运行数据，一般x i可为一多维时间序列数组。
除此之外，确定至少两个预测模型的综合准确度是否满足要求的步骤可包括：确定每个预测模型的预测准确度，当每个预测模型的预测准确度均满足要求时，确定综合准确度满足要求。
当所述至少两个预测模型的综合准确度满足要求时，可确定针对预定风力发电机组开启前馈控制功能，当所述至少两个预测模型的综合准确度不满足要求时，可确定针对预定风力发电机组不开启前馈控制功能。
以偏航控制为例，偏航控制需要根据风速和风向来确定偏航角度，以控制风力发电机组旋转确定的偏航角度，实现对风操作。当预定预测模型包括风速预定模型和风向预测模型时，根据风速预定模型所预测的风速和风向预测模型所预测的风向来确定偏航角度，以通过偏航控制方式控制预定风力发电机组旋转确定的偏航角度。
图2示出根据本申请示例性实施例的风电场中的风力发电机组的前馈控制设备的结构图。
如图2所示，根据本申请示例性实施例的风电场中的风力发电机组的前馈控制设备100包括数据获取模块10、预测模块20和前馈控制模块30。
数据获取模块10获取风电场中的预定风力发电机组的实时运行数据。获取的实时运行数据可包括预定风力发电机组运行时的实时风参数和该预定风力发电机组自身的实时运行参数。作为示例，获取的实时运行数据可为预定风力发电机组基于时间序列的与预定预测模型相应的实时运行数据。
预测模块20将获取的实时运行数据输入到与预定风力发电机组对应的预定预测模型，以通过预定预测模型获取预测数据。作为示例，通过预定预测模型所获取的预测数据可包括风参数和预定风力发电机组自身的运行参数。
预测模块20通过预定预测模型还可获取预定时间之后的预测数据，所述预定时间可为控制预定风力发电机组完成与前馈控制方式对应的动作所需的最短时间的预定倍数。
前馈控制模块30根据获取的预测数据确定是否针对预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制所述预定风力发电机组的运行状态。
在一些实施例中，风电场中的风力发电机组的前馈控制设备可还包括：测试模块(图中未示出)，判断预定预测模型的预测准确度是否满足要求。
作为示例，测试模块将获取的预定采样周期内(可指当前时刻之前的预定采样周期内)的实时运行数据输入到预定预测模型，以获得所述预定采样周期内的多个预测数据，判断所述多个预测数据和与所述多个预测数据对应的多个实际测量数据是否一致，如果所述多个预测数据与所述多个实际测量数据一致，则确定预定预测模型的预测准确度满足要求。如果预定预测模型的预测准确度满足要求，则前馈控制模块30确定针对预定风力发电机组开启前馈控制功能，如果预定预测模型的预测准确度不满足要求，则前馈控制模 块30确定针对预定风力发电机组不开启前馈控制功能。
如果所述多个预测数据与所述多个实际测量数据不一致，则表明预定预测模型的预测准确度不满足要求，此时测试模块可基于获取的实时运行数据对该预定预测模型进行在线训练，并在训练之后继续判断预定预测模型的预测准确度是否满足要求，以在预定预测模型的预测准确度满足要求之后将预定预测模型应用到前馈控制中。
作为示例，测试模块可计算预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据的平均绝对误差或平均绝对误差百分比。如果平均绝对误差大于与预定预测模型对应的设定阈值或平均绝对误差百分比大于与预定预测模型对应的设定百分比阈值，则测试模块确定所述多个预测数据与对应的所述多个实际测量数据一致，即，表明预定预测模型的预测准确度满足要求。如果平均绝对误差不大于与预定预测模型对应的设定阈值或平均绝对误差百分比不大于与预定预测模型对应的设定百分比阈值，则确定所述多个预测数据与对应的所述多个实际测量数据不一致，即，表明预定预测模型的预测准确度不满足要求。
在一些实施例中，上述预定预测模型可基于获取的预定风力发电机组的实时运行数据来进行在线训练，或者，还可基于预定风力发电机组的历史运行数据对预定预测模型进行离线训练。
作为示例，预定预测模型可包括以下至少一项：风速预测模型、风向预测模型、湍流强度预测模型、发电机转速预测模型。
第一种情况，预定预测模型可为一个预测模型，此时前馈控制模块30可根据该一个预测模型所预测的预测数据来确定是否针对预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制该预定风力发电机组的运行状态。
前馈控制模块30判断预定时间段内的预测数据变化量是否大于设定值，如果预测数据变化量大于设定值，则前馈控制模块30确定针对预定风力发电机组开启前馈控制功能，以便通过变桨控制方式或电磁扭矩控制方式控制预定风力发电机组的运行状态。
本申请示例性实施例还提供一种风电场中的风力发电机组的前馈控制系统，所述前馈控制系统除包括图2所示的前馈控制设备100之外，可还包括多个机组控制器200。每个机组控制器200可控制相应的风力发电机组的运 行状态。在此情况下，当前馈控制模块30确定针对预定风力发电机组开启前馈控制功能时，与该预定风力发电机组对应的机组控制器200可通过变桨控制方式或电磁扭矩控制方式控制预定风力发电机组的运行状态。
在第一示例中，以预定预测模型为风速预测模型为例，此时，数据获取模块10获取的与风速预测模型相应的实时运行数据可包括：实时测量的风速、风向、发电机转速、输出功率、发电机的电磁扭矩，相应地预测模型所预测的预测数据可为预测风速。
下面以预定预测模型为风速预测模型，预测数据为预测风速为例，参照图2介绍针对风速预测模型的前馈控制方式的过程。
图3示出根据本申请示例性实施例的针对风速预测模型的前馈控制过程的结构示意图。
如图3所示，假设风电场中包括n个风力发电机组，每个风力发电机组具有一个机组控制器，且一个风力发电机组对应一个风速预测模型。在此情况下，上述本申请示例性实施例所示的图2中的前馈控制设备100可包括与各风力发电机组对应的数据获取模块、与各预测模型分别对应的预测模块和测试模块，以及前馈控制模块。
具体说来，以风力发电机组1为例，针对风速预测模型的前馈控制过程可为：数据获取模块1获取风力发电机组1运行时的实时风速，预测模块1将获取的风力发电机组1运行时的实时风速输入到与风力发电机组1对应的风速预测模型1，获得风力发电机组1的预测风速。测试模块1判断风速预测模型1的预测准确度是否满足要求，当风速预测模型1的预测准确度满足要求时，则测试模块1向前馈控制模块发出使能信号，前馈控制模块响应于该使能信号，确定针对风力发电机组1开启前馈控制功能，并根据风速预测模型1所预测的预测风速提前给机组控制器1发出控制指令，机组控制器1响应于控制指令以调整电磁扭矩或者提前变桨。
如果测试模块1确定风速预测模型1的预测准确度不满足要求，则测试模块1不向前馈控制模块发送使能信号，此时前馈控制模块确定针对风力发电机组1不开启前馈控制功能，表明此时的风速预测模型1需要进一步学习以适应当前变化的风况。在此情况下，可实时获取风力发电机组1的实时运行数据，实时在线训练与风力发电机组1对应的风速预测模块1，测试模块1进行实时测试，当测试模块1确定风速预测模型1的预测准确度满足要求时， 使能前馈控制模块。
由于风速预测模型是针对各个风力发电机组建立的，即，考虑到了不同风力发电机组之间的差异，前馈控制模块可向各机组控制器分别推送不同的控制参数，以避免各风力发电机组发生过速故障或者通过提前变桨来降低风力发电机组的载荷，使得风力发电机组的效率更高、载荷更小。
前馈控制模块30可确定预定时间内的预测风速的变化量是否大于设定值(例如，设定风速变化量)。如果预测风速的变化量(可指预测风速增加的变化量或预测风速减小的变化量)大于设定值，则与预定风力发电机组对应的机组控制器可通过变桨控制方式或电磁扭矩控制方式控制预定风力发电机组的运行状态。
具体说来，如果预测风速的变化量大于设定值，则前馈控制模块30可基于预测风速确定预定风力发电机组在所述预定时间之后是否处于满发阶段。这里，前馈控制模块30可基于预测风速与额定风速的比较结果来确定预定风力发电机组是否处于满发阶段，当预测风速不小于额定风速时，前馈控制模块30可认为预定风力发电机组处于满发阶段，当预测风速小于额定风速时，前馈控制模块30可认为预定风力发电机组处于不满发阶段。
如果预定风力发电机组处于满发阶段，则前馈控制模块30可确定针对预定风力发电机组开启前馈控制功能，并确定此时的前馈控制方式为变桨控制方式，与预定风力发电机组对应的机组控制器可通过变桨控制方式控制预定风力发电机组的运行状态。当预测风速大于额定风速时，可认为此时预定风力发电机组处于满发阶段，在此情况下，与预定风力发电机组对应的机组控制器可通过变桨控制方式进行恒功率调整，即，通过变桨使预定风力发电机组恒功率输出(使风力发电机组保持以额定功率输出)。当预测风速增加的变化量大于设定风速变化量时，如果预定风力发电机组处于满发阶段，则与预定风力发电机组对应的机组控制器可控制预定风力发电机组增大桨距角，当预测风速减小的变化量大于设定风速变化量时，如果预定风力发电机组处于满发阶段，则与预定风力发电机组对应的机组控制器可控制预定风力发电机组减小桨距角。
如果预定风力发电机组处于不满发阶段，则前馈控制模块30可确定针对预定风力发电机组开启前馈控制功能，并确定此时的前馈控制方式为电磁扭矩控制方式，与预定风力发电机组对应的机组控制器可通过电磁扭矩控制方 式控制预定风力发电机组的运行状态。这里，当预测风速不大于额定风速时，可认为此时预定风力发电机组处于不满发阶段，即，处于最大风能捕获阶段，在此情况下，叶片桨距角保持在最优桨距角，与预定风力发电机组对应的机组控制器可通过电磁扭矩控制方式调整发电机电磁扭矩(如增大电磁扭矩)，调整发电机转速，使叶片运行在最优叶尖速比处，叶片的风能利用系数最大，以实现预定风力发电机组对风能的最大捕获。例如，当预测风速增加的变化量大于设定风速变化量时，如果预定风力发电机组处于不满发阶段，则与预定风力发电机组对应的机组控制器可控制预定风力发电机组增大电磁扭矩，当预测风速减小的变化量大于设定风速变化量时，如果预定风力发电机组处于不满发阶段，则与预定风力发电机组对应的机组控制器可控制预定风力发电机组减小电磁扭矩。
在第二示例中，以预定预测模型为风向预测模型为例，数据获取模块10获取的与风向预测模型相应的实时运行数据可包括：实时测量的风速、风向、机舱位置、发电机转速、输出功率、发电机的电磁扭矩。
在第三示例中，以预定预测模型为湍流强度预测模型为例，此时，数据获取模块10获取的与湍流强度预测模型相应的实时运行数据可包括：实时测量的风速、风向、发电机转速、输出功率、发电机的电磁扭矩。
在本申请实施例中，湍流强度为预定时间段的风速标准差与该预定时间段的平均风速的比值，也就是说，湍流强度为预定时间段的估计值，湍流强度预测模型所预测的湍流强度可以不作为前馈控制使用，而仅将预测的湍流强度提供给风电场的操作人员，以供操作人员了解湍流强度的变化趋势。
在此情况下，根据本申请示例性实施例的风电场中的风力发电机组的前馈控制系统还可包括：显示器，用于显示利用预定预测模型获得的预测数据。前馈控制模块可将湍流强度预测模型所预测的湍流强度发送到显示器以进行显示。
在第四示例中，预定预测模型为发电机转速预测模型，此时，数据获取模块10获取的与发电机转速预测模型相应的实时运行数据可包括：实时测量的风速、风向、发电机转速、输出功率、发电机的电磁扭矩、风力发电机组在第一预定方向(X方向)和第二预定方向(Y方向)上的加速度。
由于当风电场中的风力发电机组发生发电机过速故障时，一般无法通过远程控制的方式对故障进行复位，需要风电场的操作人员在风力发电机组的 运行现场进行操作才能复位故障。这会使得恢复上述发电机过速故障所需的故障处理时间较长，造成发电量的损失。为此，在本申请示例性实施例中可通过发电机转速预测模型来预测发电机转速，与发电机转速预测模型对应的风力发电机组对应的机组控制器基于预测的发电机转速通过变桨控制方式来提前进行调整，以避免发电机过速故障的发生，减少发电量的损失。
第二种情况，预定预测模型可包括至少两个预测模型。下面参照图4介绍针对至少两个预测模型的前馈控制方式的过程。
图4示出根据本申请示例性实施例的针对至少两个预测模型的前馈控制过程的结构示意图。
如图4所示，假设风电场中包括n个风力发电机组，每个风力发电机组具有一个机组控制器，且一个风力发电机组对应一个风速预测模型。图2中的前馈控制设备100可包括与各风力发电机组对应的数据获取模块、与各预测模型分别对应的预测模块、测试模块和调度模块，以及前馈控制模块。
具体说来，测试模块可用于确定与各风力发电机组对应的预测模型的预测准确度是否满足要求。
调度模块可用于确定至少两个预测模型的综合准确度是否满足要求。
在一个示例中，调度模块可为所述至少两个预测模型中的每个预测模型设置权重值，基于设置的权重值和每个预测模型的预测准确度确定综合准确度，当综合准确度大于预设阈值时，确定综合准确度满足要求，当综合准确度不大于预设阈值时，确定综合准确度不满足要求。
在另一示例中，调度模块当每个预测模型的预测准确度均满足要求时，确定综合准确度满足要求。
当调度模块确定所述至少两个预测模型的综合准确度满足要求时，调度模块向前馈控制模块发出使能信号，前馈控制模块响应于该使能信号，确定针对预定风力发电机组开启前馈控制功能。
如果调度模块确定所述至少两个预测模型的综合准确度不满足要求，则调度模块不向前馈控制模块发送使能信号，此时前馈控制模块确定针对预定风力发电机组不开启前馈控制功能，表明此时的所述至少两个预测模型需要进一步学习。
以偏航控制为例，偏航控制需要根据风速和风向来确定偏航角度，以控制预定风力发电机组旋转确定的偏航角度，实现对风操作。当预定预测模型 包括风速预定模型和风向预测模型时，如果风速预定模型和风向预测模型的综合准确度满足要求，则前馈控制模块根据风速预定模型所预测的风速和风向预测模型所预测的风向来确定偏航角度，与预定风力发电机组对应的机组控制器通过偏航控制方式控制预定风力发电机组旋转确定的偏航角度。
根据本申请的示例性实施例还提供一种存储有计算机程序的计算机可读存储介质。该计算机可读存储介质存储有当被处理器执行时使得处理器执行上述风电场中的风力发电机组的前馈控制方法的计算机程序。该计算机可读记录介质是可存储由计算机系统读出的数据的任意数据存储装置。计算机可读记录介质的示例包括：只读存储器、随机存取存储器、只读光盘、磁带、软盘、光数据存储装置和载波(诸如经有线或无线传输路径通过互联网的数据传输)。
根据本申请的示例性实施例还提供一种场群控制器。该场群控制器包括处理器和存储器。存储器用于存储计算机程序。所述计算机程序被处理器执行使得处理器执行如上所述的风电场中的风力发电机组的前馈控制方法。
采用本申请示例性实施例的风电场中的风力发电机组的前馈控制方法和设备，能够通过预测风力发电机组在未来时刻的运行状态以提前施加控制，从而避免极端工况对风力发电机组安全运行和载荷的影响。
此外，采用本申请示例性实施例的风电场中的风力发电机组的前馈控制方法和设备，通过将预测数据与前馈控制方式相结合，使得机组控制系统在一定程度上从被动控制系统变成了主动控制系统，并使得机组控制系统能够提前感知风速、风向、发电机转速、湍流强度的变化，并提前产生相应地控制指令，消除了目前机组被动系统带来的控制的滞后性，是实现单个风力发电机组智能化控制的基础。
此外，采用本申请示例性实施例的风电场中的风力发电机组的前馈控制方法和设备，能够提高风力发电机组出力，避免风力发电机组发生过速或加速度超限故障。
此外，采用本申请示例性实施例的风电场中的风力发电机组的前馈控制方法和设备，通过预测大湍流情况，提前给风力发电机组发出控制指令(如增大电磁扭矩)，使电磁扭矩与气动扭矩平衡，避免风力发电机组转速突增，并在避免过速类故障的同时提高了风力发电机组的出力，降低了风力发电机组的载荷。
此外，采用本申请示例性实施例的风电场中的风力发电机组的前馈控制方法和设备，在风力发电机组处于满发阶段时，通过利用风速预测模块预测出风速突然增大的情况，以通过提前变桨来卸载风速突增施加在风力发电机组上的载荷，实现风力发电机组恒功率控制。
此外，目前国内采用的风速功率预测系统的预测时间为10-20min，即每10-20min预测一次风电场的风速和功率值，以用于指导电网的调度。此时，可利用本申请示例性实施例的上述预定预测模型进行较长时间的预测，如可预测5-20min范围内的风速和功率值，以将预测的风速和功率值作为目前风速功率预测系统的重要补充。
尽管已经参照其示例性实施例具体显示和描述了本申请，但是本领域的技术人员应该理解，在不脱离权利要求所限定的本申请的精神和范围的情况下，可以对其进行形式和细节上的各种改变。

Claims (20)

  1. 一种风电场中的风力发电机组的前馈控制方法，其特征在于，所述前馈控制方法包括：
    获取风电场中的预定风力发电机组的实时运行数据；
    将获取的实时运行数据输入到与所述预定风力发电机组对应的预定预测模型，以通过所述预定预测模型获取预测数据；
    根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制所述预定风力发电机组的运行状态。
  2. 如权利要求1所述的前馈控制方法，其特征在于，根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，包括：判断所述预定预测模型的预测准确度是否满足要求，
    其中，当满足要求时，确定针对所述预定风力发电机组开启前馈控制功能。
  3. 如权利要求2所述的前馈控制方法，其特征在于，判断所述预定预测模型的预测准确度是否满足要求，包括：
    判断预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据是否一致；
    如果一致，则确定所述预定预测模型的预测准确度满足要求。
  4. 如权利要求3所述的前馈控制方法，其特征在于，判断预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据是否一致，包括：
    计算预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据的平均绝对误差或平均绝对误差百分比；
    如果所述平均绝对误差大于与所述预定预测模型对应的设定阈值或所述平均绝对误差百分比大于与所述预定预测模型对应的设定百分比阈值，则确定所述多个预测数据与对应的所述多个实际测量数据一致。
  5. 如权利要求1所述的前馈控制方法，其特征在于，根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，包括：
    判断预定时间段内的预测数据变化量是否大于设定值，
    如果大于，则通过变桨控制方式或电磁扭矩控制方式控制所述预定风力发电机组的运行状态。
  6. 如权利要求1所述的前馈控制方法，其特征在于，所述预定预测模型包括以下项中的至少一项：风速预测模型、风向预测模型、湍流强度预测模型、发电机转速预测模型。
  7. 如权利要求6所述的前馈控制方法，其特征在于，当所述预定预测模型包括至少两个预测模型时，所述前馈控制方法还包括：
    确定所述至少两个预测模型的综合准确度是否满足要求，
    其中，如果综合准确度满足要求，则确定针对所述预定风力发电机组开启前馈控制功能。
  8. 如权利要求7所述的前馈控制方法，其特征在于，确定所述至少两个预测模型的综合准确度是否满足要求，包括：
    为所述至少两个预测模型中的每个预测模型设置权重值，基于设置的权重值和每个预测模型的预测准确度确定综合准确度，当综合准确度大于预设阈值时，确定综合准确度满足要求，
    或者，当每个预测模型的预测准确度均满足要求时，确定综合准确度满足要求。
  9. 如权利要求1所述的前馈控制方法，其特征在于，通过所述预定预测模型获取预定时间之后的预测数据，其中，所述预定时间为控制所述预定风力发电机组完成与所述预定的前馈控制方式对应的动作所需的最短时间的预定倍数。
  10. 一种风电场中的风力发电机组的前馈控制设备，其特征在于，所述前馈控制设备包括：
    数据获取模块，获取风电场中的预定风力发电机组的实时运行数据；
    预测模块，将获取的实时运行数据输入到与所述预定风力发电机组对应的预定预测模型，以通过所述预定预测模型获取预测数据；
    前馈控制模块，根据所述获取的预测数据确定是否针对所述预定风力发电机组开启前馈控制功能，以便基于预定的前馈控制方式控制所述预定风力发电机组的运行状态。
  11. 如权利要求10所述的前馈控制设备，其特征在于，所述前馈控制设备还包括：测试模块，判断所述预定预测模型的预测准确度是否满足要求，
    其中，当满足要求时，前馈控制模块确定针对所述预定风力发电机组开启前馈控制功能。
  12. 如权利要求11所述的前馈控制设备，其特征在于，测试模块判断预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据是否一致，如果一致，则确定所述预定预测模型的预测准确度满足要求。
  13. 如权利要求12所述的前馈控制设备，其特征在于，测试模块计算预定采样周期内的多个预测数据和分别与所述多个预测数据对应的多个实际测量数据的平均绝对误差或平均绝对误差百分比，如果所述平均绝对误差大于与所述预定预测模型对应的设定阈值或所述平均绝对误差百分比大于与所述预定预测模型对应的设定百分比阈值，则确定所述多个预测数据与对应的所述多个实际测量数据一致。
  14. 如权利要求10所述的前馈控制设备，其特征在于，前馈控制模块判断预定时间段内的预测数据变化量是否大于设定值，如果大于，则前馈控制模块确定针对所述预定风力发电机组开启前馈控制功能，以便通过变桨控制方式或电磁扭矩控制方式控制所述预定风力发电机组的运行状态。
  15. 如权利要求10所述的前馈控制设备，其特征在于，所述预定预测模型包括以下项中的至少一项：风速预测模型、风向预测模型、湍流强度预测模型、发电机转速预测模型。
  16. 如权利要求15所述的前馈控制设备，其特征在于，所述前馈控制设备还包括：调度模块，当所述预定预测模型包括至少两个预测模型时，调度模块确定所述至少两个预测模型的综合准确度是否满足要求，
    其中，如果综合准确度满足要求，则前馈控制模块确定针对所述预定风力发电机组开启前馈控制功能。
  17. 如权利要求16所述的前馈控制设备，其特征在于，调度模块为所述至少两个预测模型中的每个预测模型设置权重值，基于设置的权重值和每个预测模型的预测准确度确定综合准确度，当综合准确度大于预设阈值时，确定综合准确度满足要求，
    或者，当每个预测模型的预测准确度均满足要求时，调度模块确定综合准确度满足要求。
  18. 一种风电场中的风力发电机组的前馈控制系统，其特征在于，所述前馈控制系统包括如权利要求10-18中的任意一项所述的风电场中的风力发电机组的前馈控制设备。
  19. 一种存储有计算机程序的计算机可读存储介质，当所述计算机程序在被处理器执行时实现如权利要求1-9中的任意一项所述的风电场中的风力发电机组的前馈控制方法。
  20. 一种场群控制器，其特征在于，所述场群控制器包括：
    处理器；
    存储器，存储有计算机程序，当所述计算机程序被处理器执行时，实现如权利要求1-9中的任意一项所述的风电场中的风力发电机组的前馈控制方法。
PCT/CN2018/100245 2018-02-28 2018-08-13 风电场中的风力发电机组的前馈控制方法和设备 WO2019165759A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
AU2018411224A AU2018411224B2 (en) 2018-02-28 2018-08-13 Feedforward control method and device for wind turbine set in wind farm
US16/640,295 US11293400B2 (en) 2018-02-28 2018-08-13 Feedforward control method and device for wind turbine set in wind farm
EP18907660.7A EP3608536B1 (en) 2018-02-28 2018-08-13 Feedforward control method and device for wind turbine set in wind farm
ES18907660T ES2962474T3 (es) 2018-02-28 2018-08-13 Método y dispositivo de control predictivo para un conjunto de aerogeneradores en un parque eólico

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810165580.8A CN110206685B (zh) 2018-02-28 2018-02-28 风电场中的风力发电机组的前馈控制方法和设备
CN201810165580.8 2018-02-28

Publications (1)

Publication Number Publication Date
WO2019165759A1 true WO2019165759A1 (zh) 2019-09-06

Family

ID=67778546

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/100245 WO2019165759A1 (zh) 2018-02-28 2018-08-13 风电场中的风力发电机组的前馈控制方法和设备

Country Status (6)

Country Link
US (1) US11293400B2 (zh)
EP (1) EP3608536B1 (zh)
CN (1) CN110206685B (zh)
AU (1) AU2018411224B2 (zh)
ES (1) ES2962474T3 (zh)
WO (1) WO2019165759A1 (zh)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11060504B1 (en) * 2020-02-07 2021-07-13 General Electric Company Systems and methods for continuous machine learning based control of wind turbines
CN111259570B (zh) * 2020-02-26 2023-10-27 沈阳工业大学 一种基于信息物理系统的风电场有功备用量化方法
CN113309663B (zh) * 2021-03-08 2022-11-15 新疆金风科技股份有限公司 用于风力发电机组的控制方法及控制装置
CN113107785B (zh) * 2021-05-12 2022-05-31 浙江浙能技术研究院有限公司 一种风电机组功率性能异常的实时监测方法及装置
US11649804B2 (en) 2021-06-07 2023-05-16 General Electric Renovables Espana, S.L. Systems and methods for controlling a wind turbine
CN113469526A (zh) * 2021-06-30 2021-10-01 华能国际电力股份有限公司广西清洁能源分公司 一种适用于复杂风区的风电场感知系统
CN115653831A (zh) * 2022-11-11 2023-01-31 国家能源集团山西电力有限公司 一种风力发电机组应急偏航控制系统和方法
CN117277452B (zh) * 2023-11-23 2024-03-26 中国电力科学研究院有限公司 一种双馈风电机组快速有功控制的方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102518553A (zh) * 2012-01-05 2012-06-27 山东电力研究院 一种用于风电场群的远程实时监控系统
CN107100794A (zh) * 2017-06-14 2017-08-29 北京唐浩电力工程技术研究有限公司 一种风轮机能量感知及控制方法
CN107524572A (zh) * 2016-06-22 2017-12-29 南京理工大学 一种风力发电机组在线状态监测和故障诊断方法
CN107559144A (zh) * 2016-06-30 2018-01-09 通用电气公司 用于风力涡轮机的前馈控制的方法及系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2005108784A1 (ja) * 2004-05-07 2005-11-17 Mitsubishi Denki Kabushiki Kaisha 風力発電評価システム、風力発電機のための予測制御サービスシステム
GB2476316B (en) * 2009-12-21 2014-07-16 Vestas Wind Sys As A wind turbine having a control method and controller for predictive control of a wind turbine generator
JP2018017198A (ja) 2016-07-29 2018-02-01 株式会社日立製作所 風力発電システム

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102518553A (zh) * 2012-01-05 2012-06-27 山东电力研究院 一种用于风电场群的远程实时监控系统
CN107524572A (zh) * 2016-06-22 2017-12-29 南京理工大学 一种风力发电机组在线状态监测和故障诊断方法
CN107559144A (zh) * 2016-06-30 2018-01-09 通用电气公司 用于风力涡轮机的前馈控制的方法及系统
CN107100794A (zh) * 2017-06-14 2017-08-29 北京唐浩电力工程技术研究有限公司 一种风轮机能量感知及控制方法

Also Published As

Publication number Publication date
AU2018411224B2 (en) 2020-12-10
US20200291921A1 (en) 2020-09-17
US11293400B2 (en) 2022-04-05
ES2962474T3 (es) 2024-03-19
EP3608536A1 (en) 2020-02-12
AU2018411224A1 (en) 2019-12-12
EP3608536C0 (en) 2023-09-27
CN110206685A (zh) 2019-09-06
CN110206685B (zh) 2020-07-28
EP3608536A4 (en) 2020-05-20
EP3608536B1 (en) 2023-09-27

Similar Documents

Publication Publication Date Title
WO2019165759A1 (zh) 风电场中的风力发电机组的前馈控制方法和设备
US9784241B2 (en) System and method for controlling a wind turbine
JP5448822B2 (ja) 風力タービンの制御システム
US10683844B2 (en) Control of a wind turbine taking fatigue measure into account
EP3613982B1 (en) Method for controlling operation of a wind turbine
US11713747B2 (en) Wind turbine control method
US10247170B2 (en) System and method for controlling a dynamic system
CN110318947B (zh) 风力发电机组的偏航控制方法、设备及系统
US20170022974A1 (en) Operating wind turbines
JP2015117682A (ja) 風力発電装置のモニタリングシステム及びモニタリング方法
US20170175709A1 (en) System and method for controlling wind turbines
CN113994087B (zh) 经由机器学习通过选择控制器来控制风力涡轮机的量的方法和系统
CN109416523A (zh) 在低电压电网事件期间使用mpc控制风力涡轮机
EP3899250B1 (en) Control of side-side and fore-aft vibrational movement of a wind turbine
EP3812577A1 (en) System and method for mitigating loads acting on a rotor blade of a wind turbine
DK201570721A1 (en) CONTROL OF A WIND TURBINE TAKING FATIGUE MEASURE INTO ACCOUNT
CN117905638A (zh) 基于强化学习的风力发电机优化控制方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18907660

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2018907660

Country of ref document: EP

Effective date: 20191106

ENP Entry into the national phase

Ref document number: 2018411224

Country of ref document: AU

Date of ref document: 20180813

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE