JP2018109369A - Wind farm and operational method thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To optimize the power generation output of the existing wind farm without changing the function of a local control device.SOLUTION: A wind farm comprises: a plurality of wind turbines; a plurality of local control devices respectively provided for the plurality of wind turbines for controlling at least one of over speed or over power of each wind turbine, and control of cut-out control of each wind turbine; and a wind farm control device which performs optimization processing of the plurality of wind turbines and is configured to give at least one of the limit value related to the pitch angle or output of each wind turbine and the command value related to the azimuth angle or wind direction deviation to each local control device based on the result of the optimization processing.SELECTED DRAWING: Figure 14

Description

  The present invention relates to a wind farm and an operating method thereof.

  Conventionally, in a wind farm that generates power with a plurality of wind turbine generators (hereinafter referred to as wind turbines), the rear wind turbine disposed on the leeward side is usually the wake of the front wind turbine relative to the front wind turbine disposed on the windward side. That is, the power generation output is smaller than that of the front windmill due to the influence of the wake. For this reason, various operation methods and control methods have been proposed in order to optimize the output of the entire wind farm, which is the total power generation output generated in each windmill.

  For example, Patent Document 1 discloses an invention related to output optimization of a wind farm. Specifically, in the technique of Patent Document 1, the wake mode of each wind turbine is defined for each wind direction in the central control device, the corrected pitch angle is determined and given to the local control device of each wind turbine, and is determined by the central control device for each wind turbine. Then, control is performed according to the given corrected pitch angle.

US Patent Publication No. 2016/0230741

  By the way, Patent Document 1 describes that a corrected pitch angle is given to the local control device. However, if an inappropriate blade pitch angle is commanded from the central control device to the local control device, there is an abnormality in the wind turbine. Since it may be generated, it is not desirable from the viewpoint of the safety of the windmill to give a command related to the blade pitch angle from the central control device (wind farm control device) to the local control device. Moreover, the said patent document 1 is not disclosed at all about the technique of installing additionally the control apparatus which optimizes the electric power generation output of this wind farm with respect to the existing wind farm.

  In view of the above-described problems, at least one embodiment of the present invention aims to optimize the power generation output of an existing wind farm without changing the function of the local control device.

(1) A wind farm according to at least one embodiment of the present invention includes:
Multiple windmills,
A plurality of wind turbines each provided for each of the plurality of wind turbines to prevent at least one of overspeed or over power of each of the wind turbines, or to perform at least one control of cutout control of each of the wind turbines. A local control unit;
Performing optimization processing of the plurality of wind turbines, and at least one of a limit value related to the pitch angle or output of each wind turbine or a command value related to azimuth angle or wind direction deviation based on the result of the optimization processing A wind farm controller configured to provide to the local controller of
Is provided.

  According to the configuration of (1) above, the control for preventing at least one of overspeed or overpower of each windmill or the cut of each windmill by the plurality of local control devices respectively provided in association with the plurality of windmills At least one of the out controls is performed. Further, the wind farm control device performs optimization processing of a plurality of wind turbines, and based on the result of the optimization processing, a limit value related to the pitch angle or output of each wind turbine, or a command value related to azimuth angle or wind direction deviation Is provided to each local controller. That is, at least one of the limit value related to the pitch angle or output of each windmill or the command value related to the azimuth angle or wind direction deviation is given from the wind farm control device to each local control device. Thereby, it is possible to additionally install a wind farm control device that optimizes the power generation output without changing the function of each local control device with respect to the existing wind farm. Therefore, the total power generation output of the wind farm can be improved through the optimization process of each wind turbine without affecting the safety of each wind turbine in the existing wind farm.

(2) In some embodiments, in the wind farm described in (1) above,
A central control unit for collecting information from at least the plurality of local control devices;
A network hub provided between the central control unit and the plurality of local control devices;
Further comprising
The wind farm control device is connected to the plurality of local control devices via the network hub.

  According to the configuration of (2) above, the wind farm control device is connected to each of the plurality of local control devices via the network hub as a control device different from the central control unit that collects information from the plurality of local control devices. Connected. That is, without changing the functions of a plurality of local control devices, a central control unit, and a network hub that centrally relays these connections corresponding to a plurality of wind turbines in an existing wind farm, the wind farm control device is connected to the network hub. It is possible to optimize the operation of an existing wind farm with a simple configuration of additionally installing.

(3) In some embodiments, in the wind farm described in (1) or (2) above,
Each of the local control devices is configured to perform a primary processing of operation data of each of the windmills and send a result of the primary processing to the wind farm control device.

  According to the configuration of (3) above, the primary processing of the operation data of each wind turbine is performed by each local control device, and the result of the primary processing is sent to the wind farm control device. Therefore, it is not necessary to perform primary processing of the operation data of each windmill in a wind farm control device connected to a plurality of local control devices. Thereby, since the calculation load in a wind farm control apparatus can be reduced, an optimization process can be performed smoothly.

(4) In some embodiments, in the wind farm according to any one of (1) to (3) above,
The wind farm control device
A setting change unit for changing a control setting of at least one setting change wind turbine among the wind turbines to be optimized;
The influence of the change in the control setting of the setting change wind turbine reaches the leeward wind turbine for each of one or more leeward wind turbines located downstream of the setting change wind turbine among the wind turbines to be optimized. A delay time calculation unit for calculating the delay time until
An evaluation value calculation unit for obtaining an evaluation value including at least the total output of the wind turbine to be optimized;
A setting update unit for updating control settings for one or more wind turbines among the wind turbines to be optimized based on the evaluation value;
Including
The evaluation value calculation unit calculates the total output of the wind turbine to be optimized using the output of each of the leeward wind turbines after the delay time has elapsed since the change of the control setting of the setting change wind turbine. The evaluation value is obtained from the calculation result of the total output.

In calculating the optimization target (evaluation function value) when optimizing the operation of the wind farm, it is desirable to consider the time delay obtained from the separation distance between the wind turbines and the wind speed.
According to the configuration of the above (4), when the control setting of the setting change wind turbine among the optimization target wind turbines is changed by the setting change unit, each of the one or more leeward wind turbines located on the downstream side of the setting change wind turbine. , The delay time until the influence of the setting change applied to the setting change wind turbine reaches each leeward wind turbine is calculated by the delay time calculation unit. Further, the evaluation value calculation unit calculates the total output of the wind turbine to be optimized using the outputs of the leeward wind turbines after the delay time has elapsed from the change of the control setting of the setting change wind turbine, and the total output An evaluation value is obtained based on the calculation result. And based on the evaluation value calculated | required by the evaluation value calculation part, a setting update part updates a control setting about one or more windmills among the optimization target windmills. Therefore, the control setting of the downstream wind turbine is changed based on the evaluation value obtained in consideration of the output of the leeward wind turbine at the timing when the setting change applied to the upstream wind turbine reaches the downstream wind turbine. Therefore, it is possible to obtain a more appropriate evaluation value in accordance with the actual situation of the wind farm. This makes it possible to set an optimization target for the total power output of the entire wind farm while appropriately considering the delay time of the influence on the downstream wind turbine due to the setting change applied to the upstream wind turbine. Furthermore, by additionally applying such a control device to an existing wind farm, the output optimization of the wind farm can be effectively realized with a small-scale modification.

(5) In some embodiments, in the wind farm described in (4) above,
The wind farm control device includes a data processing start command unit for sending a data processing start command to the plurality of local control devices so as to start primary processing of operation data including outputs of the plurality of wind turbines,
Each of the local control devices is configured to perform the primary processing of the operation data of each of the wind turbines and send a result of the primary processing to the wind farm control device.

  According to the configuration of (5) above, each local control device performs primary processing of operation data including outputs of a plurality of wind turbines according to the data processing start command transmitted by the data processing start command unit, and the primary processing The result of the processing is transmitted from each local control device to the wind farm control device. Thereby, the operator of the wind farm control device can collect the operation data of each wind turbine at a necessary timing as necessary, and, for example, based on the collected result of the primary processing, for example, restrictions on the pitch angle or the output. At least one of the value or the command value related to the azimuth angle or the wind direction deviation can be transmitted to each local control device.

(6) In some embodiments, in the wind farm described in (4) or (5) above,
The at least one setting change windmill is:
A first setting change windmill;
A second setting change wind turbine located downstream of the first setting change wind turbine;
Including
The setting change unit
Until the influence of the change of the control setting of the first setting change wind turbine reaches the second setting change wind turbine, the previous value of the control setting of the second setting change wind turbine is held,
When the influence of the change of the control setting of the first setting change wind turbine reaches the second setting change wind turbine, the control setting of the second setting change wind turbine is changed.

  According to the configuration of (6) above, until the influence of the change in the control setting in the first setting change wind turbine reaches the second setting change wind turbine located downstream of the first setting change wind turbine. In the 2-setting change wind turbine, the previous control set value is held. That is, the control setting of the second setting change wind turbine is not changed before the delay time until the influence of the change of the control setting in the first setting change wind turbine reaches the second setting change wind turbine. The control setting of the second setting change wind turbine is changed after the delay time has elapsed. Accordingly, the control setting of the second setting change wind turbine can be changed at an appropriate timing with respect to the influence of the control setting change on the first setting change wind turbine at an appropriate timing at which the improvement effect due to the setting change of the second setting change wind turbine can appear. . Moreover, the possibility of the output reduction by changing the control setting of the 2nd setting change windmill before progress of delay time can be reduced. As a result, the total power generation output of the entire wind farm can be optimized.

(7) In some embodiments, in the wind farm according to any one of (4) to (6) above,
The control setting includes at least one of a power generation output command, a pitch angle command, or an azimuth command for each of the wind turbines.

  According to the configuration of (7) above, the control setting of the setting change wind turbine changed by the setting changing unit includes one or more of the power generation output command, pitch angle or direction command of each wind turbine. In other words, while appropriately considering the delay time until the influence of the change in the control setting including at least one of the power generation output command, the pitch angle command, or the direction command of each wind turbine reaches the downstream wind turbine, The total power output can be optimized.

(8) A wind farm operating method according to at least one embodiment of the present invention includes:
A wind farm operating method comprising: a plurality of wind turbines; a plurality of local control devices provided for the plurality of wind turbines; respectively; and a wind farm control device for performing optimization processing of the plurality of wind turbines. ,
Performing at least one of control of preventing at least one of overspeed or overpower of each of the wind turbines, or cutout control of each of the wind turbines, by each of the local control devices;
Performing the optimization process of the plurality of wind turbines by the wind farm control device;
Based on the result of the optimization process, at least one of a limit value related to the pitch angle or output of each windmill or a command value related to azimuth angle or wind direction deviation is sent from the wind farm control device to each local control device. Giving step,
Is provided.

  According to the configuration of (8) above, the control for preventing at least one of overspeed or overpower of each windmill or the cut of each windmill by the plurality of local control devices respectively provided in association with the plurality of windmills. At least one of the out controls is performed. Further, the wind farm control device performs optimization processing of a plurality of wind turbines, and based on the result of the optimization processing, a limit value related to the pitch angle or output of each wind turbine, or a command value related to azimuth angle or wind direction deviation Is provided to each local controller. In other words, one or more of the limit value related to the pitch angle or output of each wind turbine or the command value related to the azimuth angle or wind direction deviation is given from the wind farm control device to each local control device. Thereby, it is possible to additionally install a wind farm control device that optimizes the power generation output without changing the function of each local control device with respect to the existing wind farm. Therefore, the total power generation output of the wind farm can be improved through the optimization process of each wind turbine without affecting the safety of each wind turbine in the existing wind farm.

(9) In some embodiments, in the wind farm operating method according to (8) above,
Performing a primary process of operation data of each of the wind turbines by each of the local control devices;
Sending the result of the primary processing to the wind farm control device.

  According to the configuration of (9) above, the primary processing of the operation data of each windmill is performed by each local control device, and the result of the primary processing is sent to the wind farm control device. Therefore, it is not necessary to perform primary processing of the operation data of each windmill in a wind farm control device connected to a plurality of local control devices. Thereby, since the calculation load in a wind farm control apparatus can be reduced, an optimization process can be performed smoothly.

(10) In some embodiments, in the wind farm operating method according to (8) or (9) above,
In the step of performing the optimization process,
Change the control setting of at least one setting change wind turbine among the wind turbines to be optimized,
The influence of the change in the control setting of the setting change wind turbine reaches the leeward wind turbine for each of one or more leeward wind turbines located downstream of the setting change wind turbine among the wind turbines to be optimized. Calculate the delay time until
Obtain an evaluation value including at least the total output of the wind turbine to be optimized,
Based on the evaluation value, while updating control settings for one or more windmills among the windmills to be optimized,
When obtaining the evaluation value, the total output of the wind turbine to be optimized is calculated using the output of each of the leeward wind turbines after the delay time has elapsed since the change of the control setting of the setting change wind turbine. Then, the evaluation value is obtained from the calculation result of the total output.

  According to the configuration of (10) above, when the control setting of the setting change wind turbine among the optimization target wind turbines is changed, the setting change is made for each of the one or more leeward wind turbines located on the downstream side of the setting change wind turbine. A delay time until the influence of the setting change applied to the windmills reaches each leeward windmill is calculated. Further, the total output of the wind turbine to be optimized is calculated using the output of each leeward wind turbine after a delay time has elapsed from the time when the control setting of the setting change wind turbine is changed, and based on the calculation result of the total output An evaluation value is obtained. Then, based on the obtained evaluation value, the control setting is updated for one or more wind turbines among the wind turbines to be optimized. Therefore, the control setting of the downstream wind turbine is changed based on the evaluation value obtained in consideration of the output of the leeward wind turbine at the timing when the setting change applied to the upstream wind turbine reaches the downstream wind turbine. Therefore, it is possible to obtain a more appropriate evaluation value in accordance with the actual situation of the wind farm. This makes it possible to set an optimization target for the total power output of the entire wind farm while appropriately taking into account the influence on the downstream wind turbine due to the setting change applied to the upstream wind turbine.

(11) In some embodiments, in the wind farm operating method according to (10) above,
The wind farm control device sends a data processing start command to the plurality of local control devices so as to start primary processing of operation data including outputs of the plurality of wind turbines,
Each local control device may be configured to perform the primary processing of the operation data of each windmill and send the result of the primary processing to the wind farm control device.

  According to the configuration of (11) above, in response to the data processing start command transmitted by the wind farm control device, each local control device performs primary processing of operation data including outputs of a plurality of wind turbines, and the primary processing Is sent from each local control device to the wind farm control device. Thereby, the operator of the wind farm control device can collect the operation data of each wind turbine at a necessary timing as necessary, and, for example, based on the collected result of the primary processing, for example, restrictions on the pitch angle or the output. At least one of the value or the command value related to the azimuth angle or the wind direction deviation can be transmitted to each local control device.

  According to at least one embodiment of the present invention, it is possible to provide a renewable energy type power generator that reduces the nacelle weight and improves the assemblability. Moreover, according to at least one embodiment of the present invention, it is possible to provide a method for assembling a renewable energy type power generator with improved assemblability.

It is the schematic which shows the structural example of the wind farm which concerns on one Embodiment. It is the schematic which shows the structural example of the windmill in one Embodiment. It is a block diagram which shows the structure of the local control apparatus in one Embodiment. It is a block diagram which shows the structure of the control system in the control apparatus of the wind farm which concerns on one Embodiment. It is the schematic for demonstrating the improvement of the total electric power generation output by the control method of the wind farm in one Embodiment. It is a schematic diagram which shows the processing concept of the simultaneous perturbation probability approximation method (SPSA) which is the basic control method in one Embodiment. It is a schematic diagram which shows the processing concept of the multi-resolution simultaneous perturbation probability approximation method (MR-SPSA) which is the basic control method in one Embodiment, (a) shows an example of low-resolution mode, (b) is high An example of the resolution mode is shown. It is a figure which shows the control setting, evaluation value, and update algorithm in one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment. It is a schematic diagram for demonstrating the transmission delay of the wind in the wind farm which concerns on one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment. It is a flowchart which shows the output optimization process by the control apparatus of the wind farm which concerns on one Embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, and the like of the components described in this embodiment are not intended to limit the scope of the present invention unless otherwise specified, and are merely illustrative examples. Only.

  FIG. 1 is a diagram showing a configuration example of a wind farm. FIG. 2 is a view showing an example of a windmill constituting a wind farm.

As shown in FIG. 1, a wind farm 1 according to some embodiments includes a plurality of wind turbines T (T1 to Tn), wind farms electrically connected to the wind turbines T through signal lines 2, respectively. And a controller 10 (wind farm control device). The wind farm controller 10 controls operation of the wind farm 1 including a plurality of wind turbines T1 to Tn. In some embodiments, the wind farm 1 may include a plurality of local control devices LC (LC1 to LCn) provided for the plurality of wind turbines T, respectively. In some embodiments, the wind farm 1 includes a SCADA server 4 (central control unit) for collecting information from at least a plurality of local control devices LC, and the SCADA server 4 and the plurality of local control devices LC. And a network hub 3 provided therebetween. In some embodiments, the wind farm controller 10 may be connected to multiple local controllers LC via the network hub 3.
Hereinafter, after giving examples of the windmills T1 to T3 constituting the wind farm 1, details of the local control device LC and the wind farm controller 10 will be described. In FIG. 1, three wind turbines T1 to T3 are illustrated as a plurality of wind turbines. However, the number of wind turbines constituting the wind farm 1 is not limited to this, and an arbitrary number (for example, n or N). It may be.

  In some embodiments, each windmill T1 to T3 of the wind farm 1 is connected to the hub 104 and the rotor 105, which includes a plurality of blades 102 and a hub 104 to which they are attached, as shown in FIG. A main shaft 106 and a generator 110 driven by the rotational force of the main shaft 106 may be provided. In some embodiments, the main shaft 106 and the generator 110 may be connected via a drive train 108 and its output shaft 109. In some embodiments, the drive train 108 may include a gear-type gearbox that speeds up the rotation of the main shaft 106. In some embodiments, the drive train 108 may include a hydraulic transmission instead of a geared gearbox. In another embodiment, instead of the drive train 108, a direct drive system in which the main shaft 106 and the generator 110 are directly connected may be used.

  The drive train 108 and the generator 110 may be housed in a nacelle 112 that rotatably supports the main shaft 106 via a main shaft bearing 107. The nacelle base plate 114 constituting the bottom surface of the nacelle 112 may be supported by the tower 116 via a yaw slewing bearing 118. A yaw turning mechanism 119 having a yaw motor 21 (see FIG. 3) and a pinion gear may be fixed to the nacelle base plate 114, and a pinion gear of the yaw turning mechanism 119 is attached to a ring gear provided on the tower 116 side. The nacelle 112 may be turned with respect to the tower 116 by driving the yaw motor 21 in a state where the two are engaged. Further, each blade 102 is supported by the hub 104 via a slewing bearing (not shown), and the pitch angle can be adjusted by a pitch drive actuator 23 (see FIG. 3) provided in the hub 104. May be.

In the wind turbine T having the configuration example shown in FIG. 2, state values indicating damage or deterioration states of various parts are acquired by a state value detection sensor (for example, including the load sensor 33 shown in FIG. 3), It is reported to the farm controller 10.
Specific examples of the state values of the various components of each windmill T1 to T3 include the weight of the blade 102, the weight imbalance between the blades 102, the vibration of the blade 102, etc. as indicating the damaged or degraded state of the blade 102. As an example of the damaged or deteriorated state of the tower 116, the fatigue load at the base of the tower 116 or the upper portion of the tower 116, the vibration of the tower 116, and the like can be given. Damage to the drive train 108 such as a gearbox or a hydraulic transmission Examples of the state or deterioration state include vibration of the main shaft bearing 107 and the hydraulic pump bearing, vibration and swing of the main shaft 106, piston vibration and amplitude of the hydraulic pump, efficiency of the hydraulic transmission, and the like. That shows damage or deterioration of cast or cast member such as hub 104 Mention may be made of the fatigue due to stress concentration in.

Next, details of the local control device LC will be described. FIG. 3 is a block diagram illustrating a configuration of a control system in the local control device LC according to some embodiments.
As shown in FIG. 3, in some embodiments, the local control device LC is, for example, a computer, and a CPU 51 and a ROM as a storage unit for storing data such as various programs executed by the CPU 51 and tables. (Read Only Memory) 53, a RAM (Random Access Memory) 52 functioning as a work area such as a development area and a calculation area when executing each program, a hard disk drive (HDD) as a large capacity storage device (not shown), A communication interface for connecting to a communication network and an access unit to which an external storage device is attached may be provided. These are all connected via a bus 54, and the bus 54 is electrically connected to the wind farm controller 10 and the SCADA server 4 via the signal line 2 (see FIG. 1) and the network hub 3. Further, the wind farm controller 10 may be connected to, for example, an input unit (not shown) including a keyboard and a mouse, a display unit (not illustrated) including a liquid crystal display device that displays data, and the like.

  In some embodiments, the local control device LC is transmitted with detection signals relating to the wind direction, the wind speed, and the load from each of the wind direction sensor 31, the wind speed sensor 32, and the load sensor 33 provided in each of the wind turbines T1 to T3. May be. One or more of the load sensors 33 may be installed in a place where an apparatus load or a wind load acts, such as the spindle bearing 107 or the tower 116. In some embodiments, the wind farm controller 10 is electrically connected to the yaw motor 21, the yaw brake drive actuator 22, the pitch drive actuator 23, and the pitch brake drive actuator 24 via the bus 14 and the signal line 2. Also good.

  In some embodiments, the ROM 13 may store a safety control program 55 described later.

Next, details of the wind farm controller 10 will be described.
FIG. 4 is a block diagram illustrating a configuration of a control system in the wind farm controller 10 (wind farm control device) according to some embodiments.
As shown in FIG. 4, in some embodiments, the wind farm controller 10 is, for example, a computer and serves as a CPU 11 and a storage unit for storing data such as various programs executed by the CPU 11 and tables. ROM 13, RAM 12 that functions as a work area such as a development area and a calculation area when executing each program, a hard disk drive (HDD) as a mass storage device (not shown), a communication interface for connection to a communication network, and An access unit to which an external storage device is attached may be provided. In some embodiments, the wind farm controller 10 may include a database 19 that stores the optimal control settings obtained in the process of updating the control settings described below in association with the wind condition parameters. A power generation output distribution table (not shown) or the like may be stored. These are all connected via a bus 14, and the bus 14 is electrically connected to each wind turbine T1 to T3 of the wind farm 1 via a signal line 2 (see FIG. 1). Further, the wind farm controller 10 may be connected to, for example, an input unit (not shown) including a keyboard and a mouse, a display unit (not illustrated) including a liquid crystal display device that displays data, and the like.

  In some embodiments, the wind farm controller 10 receives detection signals related to the wind direction, wind speed, and load from the wind direction sensor 31, the wind speed sensor 32, and the load sensor 33 provided in each of the wind turbines T1 to T3. May be. In some embodiments, the wind farm controller 10 is electrically connected to the yaw motor 21, the yaw brake drive actuator 22, the pitch drive actuator 23, and the pitch brake drive actuator 24 via the bus 14 and the signal line 2. Also good.

  In some embodiments, the ROM 13 includes an output calculation program 15 that calculates the entire output of the wind farm 1 from the power generation amount of each of the wind turbines T1 to T3, and a first output for optimizing the entire output of the wind farm 1. An optimization program 16 (operation control program), a second output optimization program 17 (operation control program), and a third output optimization program 18 (operation control program) may be stored. These programs will be described later.

  Here, the output optimization process of the wind farm 1 in some embodiments will be described in detail. The optimization process of the wind farm in some embodiments described in this specification is not limited to the optimization process focusing on a specific parameter, and can be applied to various optimization processes. Specifically, for example, an optimization process focusing on the power generation output may be used. In another embodiment, for example, an optimization process for reducing mechanical fatigue of each wind turbine T may be used. Further, as a technique used for the optimization process at that time, a simultaneous perturbation probability approximation method (SPSA) which is one of the conventional optimization techniques may be used, or a multi-resolution simultaneous perturbation probability approximation method developed from this. (MR-SPSA) may be used. In some embodiments, the optimization of the output of the entire wind farm 1 can be achieved in consideration of the influence of wake by applying the further developed optimization process based on MR-SPSA.

  First, a simultaneous perturbation probability approximation method (SPSA) and a multi-resolution simultaneous perturbation probability approximation method (MR-SPSA) will be described as examples of optimization methods.

  In SPSA, control is performed in a so-called high resolution mode in which all individual wind turbines T belonging to the wind farm 1 are controlled independently and controlled using independent parameter setting values (see FIG. 5). Specifically, a calculation algorithm that determines and gives the next perturbation based on the evaluation value (evaluation function) after giving a positive or negative perturbation, which is a random variable, to each parameter, and repeats it until there is no allowance for improvement The wind turbines T are controlled according to the following.

  FIG. 6 is a schematic diagram showing a processing concept of a simultaneous perturbation probability approximation method (SPSA) which is a basic control method in some embodiments. In some embodiments, the plant 41 shown in FIG. 6 may be, for example, a wind farm 1 including a plurality of windmills T1 to Tn.

  In general, a wind power plant such as wind farm 1 is a power plant that uses renewable energy. For example, wind conditions such as wind direction, wind speed, temperature, humidity, air density, etc. change every moment. It is difficult to specify clearly.

  Therefore, in some embodiments, for example, as illustrated in FIG. 6, the plant 41 (Plant) is set as a control target whose transfer function is unknown, and an input value (Measured input) from the controller 40 to the plant 41, Feedback control that measures an output value (Measured output) from the plant 41 when the input value is input and determines the next input value depending on whether or not the output value approaches the control target 42 (Control objective). Is done. The output value from the plant 41 is also compared with a reference input value to the controller 40, and the difference (Measured error) is input to the controller 40 and the control target 42.

Then, a control result (Control performance) in each loop (hereinafter referred to as iteration) of the loop control repeatedly performed is input from the control target 42 to the optimization tool (Optimization tool) 43. The optimization tool 43 receives the control result in the first iteration and applies perturbations to a plurality of control parameters with respect to the input value to the plant 41 in the next second iteration. Perturbations first control settings (e.g., theta _k) with respect to, for example, is a random variable + delta _k or - [delta _k is given. In some embodiments, when the perturbation is given, the perturbation may be given simultaneously to all the wind turbines T to be optimized. If it does in this way, the optimal control setting as the whole wind farm 1 can be reached quickly in a short time, and the total power generation output of the whole wind farm 1 can be improved in a short time.

  Subsequently, the gradient of the evaluation value (evaluation function L) is obtained in the optimization tool 43 from the above perturbation and the control result of the second iteration given the perturbation, and the next control parameter is determined. It is sent to the controller 40 as a tuning command to the controller 40 in the next iteration. By repeating this, an optimum solution, that is, an optimized control setting is searched.

  On the other hand, in MR-SPSA, the number of all wind turbines T (for example, N) belonging to the wind farm 1 to be controlled is less than the total number (for example, M, where N> M). It is divided into groups, and the low resolution mode for controlling the wind turbines T in the same group using the same parameter setting values, and all the individual wind turbines T are set as independent control targets, and independent parameter setting values for each. Switch to the high-resolution mode controlled using, and determine the next perturbation based on the evaluation value (evaluation function) after giving positive or negative perturbation that is a random variable to each parameter. The wind turbines T are controlled according to a calculation algorithm that is repeated until the improvement cost is eliminated.

In the MR-SPSA, in each iteration of the iterative calculation for obtaining the optimum setting solution, the first perturbation component Δ _k1 is used as the perturbation component in the low resolution mode, and the second perturbation component Δ _k2 is used as the perturbation component in the high resolution mode. Is used. That is, in MR-SPSA, one of the first perturbation component Δ _k1 and the second perturbation component Δ _k2 is given as perturbation.

  In the MR-SPSA, the control mode is set to the low resolution mode in the initial stage of the iterative calculation for obtaining the optimal setting solution, and for example, rough correction (update of the control setting value) is performed for each group with respect to the yaw rotation and the pitch angle. Is called. When the evaluation value changes below a certain value, the control mode is switched to the high resolution mode, and fine adjustment of details is performed for each windmill. As a result, the time required to reach the optimum setting is shortened, and the power generation efficiency of the entire wind farm can be increased (see FIG. 5).

  FIG. 7A and FIG. 7B are control methods of the wind farm 1 in some embodiments, and processing of a multi-resolution simultaneous perturbation probability approximation method (MR-SPSA), which is a basic control method. FIG. 7A is a schematic diagram showing a concept, FIG. 7A shows an example of a low resolution mode, and FIG. 7B shows an example of a high resolution mode. As shown in FIGS. 7A and 7B, the wind farm 1 has, for example, four rows in the vertical direction (Y direction) in the figure, and the horizontal direction (X direction and X direction) perpendicular to the Y direction. In total, 16 wind turbines T1 to T16 arranged in four rows may be included. These windmills T1 to T16 may be arranged at equal intervals in the X direction and the Y direction. The distance between each windmill T may be represented by a multiple of D, where D is the direct system of the rotor 105, for example. For convenience of explanation, the 16 wind turbines in FIGS. 7A and 7B are numbered T1 to T16 in order from the upper left to the right.

Now, it is assumed that the wind is blowing from the lower left to the upper right (eg, wind direction vector (x, y) = (1, 1)).
In this case, the wind turbine T in front of the windward side, that is, upstream of the wind direction (for example, wind turbines T5 to T7, T9 to T11, and T13 to T15 in three rows and three columns from the lower left shown by a broken line in FIG. 7A). To the group G1), the rear wind turbine T arranged on the leeward side, that is, the downstream side of the wind direction (for example, the wind turbines T1 to T4 in the uppermost and rightmost rows indicated by broken lines in FIG. 7A). (T8, T12, and T16 are group G2) Usually, the power generation output of each wind turbine T is smaller than that of the wind turbine T included in the group G1 due to the influence of the wake of the front wind turbine T, that is, the wake. Therefore, in the low resolution mode shown in FIG. 7A, the wind turbines T in the group G1 located on the windward side are controlled using the same control parameter (for example, θ _k1 ) in the group G1, and the leeward Each wind turbine T in the group G2 located on the side may also be configured to be controlled using the same control parameter (for example, θ _k2 ) in the group G2.
The grouping is not limited to the groups G1 and G2 described above, and there can be various patterns. For example, in FIG. 7A, for the wind blowing from the left, the group including at least the leftmost wind turbines T1, T5, T9 and T13 is the upstream group G1, and the rightmost wind turbines T4, T8, A group including at least T12 and T16 may be set as the downstream group G2. Furthermore, it may be grouped with other patterns.

  On the other hand, in the high resolution mode shown in FIG. 7B, each of the wind turbines T (T1 to T16) is a control target, and each wind turbine T1 to T16 is controlled using individually independent control parameters. May be configured.

Based on the above, for example, processing in the low resolution mode performed by grouping into the windward group G1 and the windward group G2 with the above-described windmills T1 to T16 as control targets will be described.
In some embodiments, let the random variation in the low resolution mode be Δ _k1 and the random variation in the high resolution mode be Δ _k2 . In some embodiments, Δ _k2 is a random variable where each element is independent. On the other hand, elements of delta _k1 in the low-resolution mode, for example, the only two delta _k11 and delta _k12. This is because the wind turbine T in each group (indicated by a broken line in FIG. 7A) is controlled using the same control setting. Figure 7 perturbation delta _k1 in the wind turbine T1~T16 of (a) is represented by the following formula (1) in order.

[Equation 1]
_Δk1 = [ _{Δk11, Δk11, Δk11, Δk11, Δk12, Δk12, Δk12, Δk11, Δk12, Δk12, Δk12, Δk11, Δk12, Δk12, Δk12,} Δk12 _k11 ] ^T (1)

At this time, as shown in FIG. 8, for example, the control setting is represented by Expression (2), the evaluation value L is represented by Expression (3), and the control setting update algorithms are represented by Expressions (4) to (6), respectively. It may be set as represented by
Here, a _k1 , a _k2 , c _k1 , and c _k2 are gains, respectively, and Δ _k1 and Δ _k2 are perturbations composed of dimensionless random variables.
In other words, in some embodiments, unlike the conventional MR-SPSA using delta _k1 or delta _k2 instead of delta _k in SPSA, simultaneously optimizing the control with the delta _k1 and delta _k2 instead of delta _k in SPSA Is done.

In some embodiments, by the above-described optimization control, for example, as shown by a thick line in FIG. 5, the time until optimization of the entire output of the wind farm 1 is compared with that of the conventional SPSA or MR-SPSA. And further shortened. In some embodiments described above, since the low resolution mode and the high resolution mode are always used in combination, there is no need to consider the switching timing between the low resolution mode and the high resolution mode. This is due to the fact that control is always given an optimal perturbation to the change.
In the high resolution mode when the steady state is ensured for a long time with little wind change, it is desirable to repeatedly execute the control loop under a certain wind condition and converge until the optimal control setting is reached. In the actual wind farm 1, since the wind changes every moment, it is not always necessary to converge to the optimum control setting in the high resolution mode. Rather, the wind farm 1 as a whole can optimize the total output of power generation by always operating in a setting state close to the optimal control setting as much as possible while constantly following the ever-changing wind.

  Subsequently, with the above-described configuration, the operation method of the wind farm 1 realized in some embodiments, that is, the output optimization control realized by the wind farm controller 10 executing the first output optimization program 16. (Output optimization processing) will be described. In the following description, the processing content performed by the wind farm controller 10 executing the first output optimization program 16 is the operation method of the wind farm 1.

In some embodiments, the wind farm controller 10 of the wind farm 1 including a plurality of wind turbines T reads the first output optimization program 16 stored in the ROM 13, expands it in the RAM 12, and executes it. , for example, as shown in FIG. 9, it executes the processing perturbing delta _k to the control settings theta _k for each wind turbine T to be optimized (step S1). The 1st output optimization program 16 functions as a perturbation provision part for giving a perturbation to each control setting of the optimization target windmill T by making the wind farm controller 10 implement | achieve said process. Then, the wind farm controller 10, after giving the perturbation delta _k, executes a process for obtaining the evaluation value L including at least the total output of the wind turbine T that is optimized (step S2). The output optimization program 16 realizes the above-described processing by causing the wind farm controller 10 to perform the above processing, and then gives an evaluation value for obtaining an evaluation value L including at least the total output of the wind turbine T to be optimized. It functions as a part. The wind farm controller 10, based on the gradient of the evaluation value L, may execute processing of updating the control settings theta _k for each of the wind turbine T to be optimized (step S3). As a result, the control setting θ _k is updated to θ _{k + 1} . The output optimization program 16 updates the control setting θ _k for each of the optimization target wind turbines T based on the gradient of the evaluation value L by causing the wind farm controller 10 to perform the above processing. It functions as a part.

In some embodiments, step (step S1) perturbing delta _k above, for example, as shown in FIG. 10, for each of the M groups corresponding to the low resolution mode of the wind turbine T that belong to each group You may set 1st perturbation component (DELTA) _k1 which should be given with respect to control setting (theta) _k (step S11). Further, a second perturbation component Δ _k2 to be given to the control setting θ _{k of the} wind turbine T belonging to each group may be set for each of N (where N> M) groups corresponding to the high resolution mode. (Step S12). Then, a perturbation including at least the first perturbation component _Δk1 and the second perturbation component _Δk2 may be given to each windmill T (step S13). In some embodiments described above, for example, M = 2 (groups G1, G2) and N = 16 (total number of windmills T in the wind farm 1) may be used.

In the above configuration, each of the windmills T included in the wind farm 1 always has a perturbation including both the first perturbation component Δ _k1 corresponding to the low resolution mode and the second perturbation component Δ _k2 corresponding to the high resolution mode. Given to. Then, based on the gradient of the evaluation value L obtained after that, the control setting θ _k is updated for each of the wind turbines T to be optimized. As a result, the individual windmills T can, for example, roughly respond to changes in wind direction and wind speed in the low resolution mode that is immediately effective as a whole of the wind farm 1 and the optimum control setting θ at the position of the individual windmills T. _Since it is possible to perform in parallel without switching the detailed correspondence in the high-resolution mode that is finely corrected to _k , the entire wind farm 1 can be operated so as to always perform an optimal response. Therefore, a conventional wind farm (for example, a window using SPSA or MR-SPSA shown in FIG. 5) in which one of the first perturbation component Δ _k1 and the second perturbation component Δ _k2 is given to each wind turbine T. Compared to the farm), even when the optimum control setting θ _k of each wind turbine T changes due to the fluctuation of the wind, the new optimum control setting θ _{k + n} can be reached in a short time. In addition, since all wind turbines T are included in the grouping target in the low resolution mode and the optimization target in the high resolution mode, the wind farm 1 is compared with the conventional wind farm in which the control setting is performed one by one. Overall power generation output can be improved in a short time. Further, by additionally applying the wind farm control device 10 to the existing wind farm, the output optimization of the wind farm 1 can be effectively realized with small modifications or updates.

  In some embodiments, when performing the various optimization processes described above, the wind farm controller 10 further determines whether the wake of the windward wind turbine T is based on the separation distance D and the wind speed V between the wind turbines T. Considering a delay time t until the windward wind turbine T is reached, a process of reflecting the delay time t in the control settings of the windward wind turbine T is executed.

When considering the separation distance D and the wind speed V between the wind turbines T, installation information regarding the installation position of the wind turbine T, wind speed information and wind direction information measured by each wind turbine T, and the like are used. In some embodiments, a wind turbine characteristic coefficient representing a characteristic specific to each wind turbine T may be included.
The wind speed information used here is preferably not the inflow wind speed with respect to the wind turbine T but the wake wind speed.
As the wake wind speed, for example, the wind speed measured by the wind direction / anemometer (wind direction sensor 31 and wind speed sensor 32) provided in the nacelle 112 of the windward wind turbine T may be used. You may use the wind speed measured by the provided rider (LIDAR). In some embodiments, the inflow wind speed to the windward windmill T, the pitch angle of the windward windmill T, and the wind speed obtained by the arithmetic processing using the windmill characteristic coefficient specific to the windward windmill T are used. May be.

  Referring to FIG. 11, the wind speed flowing into windmill T1 at time t = t1 is V1. The control setting optimized based on this information is optimal for the windmill T1 at time t = t1, but there is a time delay for the effect of changing the control setting of the windmill T1 to be transmitted to the windmill T2 and beyond. It may not necessarily be optimal for the wind turbine T after T2. Therefore, it is not desirable to change the control settings of the windmills T2, T3... At time t = t1.

  In some embodiments, if the wind speed of the wind turbine T1 after changing the control setting of the wind turbine T1 to the optimized control setting is V2, the influence of the change in the control setting of the wind turbine T1 reaches the wind turbine T2. The delay time until is t = D2 / V2. Therefore, it is desirable that the control setting change of the wind turbine T2 is time t = t1 + D1 / V2. Similarly, the control setting change of the windmill T3 is desirably time t = t1 + D1 / V2 + D2 / V3, and the same applies to the windmills T4 to Tn thereafter. Further, since V2, V3,... Change from moment to moment, it is more desirable to consider this change.

  In some embodiments, when the control settings (for example, the wind speed ratio between the wind turbines T) obtained by optimization are commanded to the wind turbines T, the separation distance D and the wind speed V between the wind turbines T are taken into consideration. The rear wind turbine T maintains the previous command value until the influence of the change in the control setting of the front wind turbine T is reached, and the control setting is changed after the influence of the control setting change of the front wind turbine T is reached. Also good.

  Subsequently, with the above-described configuration, the operation method of the wind farm 1 realized in some embodiments, that is, the above-described time delay realized by the wind farm controller 10 executing the second output optimization program 17. The output optimization control (output optimization process) in consideration of the above will be described. In the following description, the processing content performed by the wind farm controller 10 executing the second output optimization program 17 is the operation method of the wind farm 1.

In some embodiments, the wind farm controller 10 reads the second output optimization program 17 stored in the ROM 13, expands it in the RAM 12, and executes it, for example, as shown in FIG. The control setting of at least one setting change wind turbine T among the optimization target wind turbines T may be changed (step S101). The second output optimization program 17 causes the wind farm controller 10 to realize the above processing, thereby changing the setting change unit for changing the control setting of at least one setting change wind turbine T among the wind turbines T to be optimized. Function as. Next, the wind farm controller 10 sets the control setting of the setting change wind turbine T for each of the one or more leeward wind turbines T located downstream of each setting change wind turbine T among the optimization target wind turbines T. A delay time until the influence of the change reaches the leeward wind turbine T may be calculated (step S102). The second output optimization program 17 causes the wind farm controller 10 to realize the above-described processing, so that one or more leeward sides located downstream of the setting change wind turbines T among the wind turbines T to be optimized. For each wind turbine T, the wind turbine T functions as a delay time calculation unit for calculating a delay time until the influence of the control setting change of the setting change wind turbine T reaches the leeward wind turbine T. Next, the wind farm controller 10 may obtain an evaluation value including at least the total output of the wind turbine T to be optimized (step S103). The second output optimization program 17 functions as an evaluation value calculation unit for obtaining the evaluation value L including at least the total output of the wind turbine T to be optimized by causing the wind farm controller 10 to realize the above processing. . Next, the wind farm controller 10 may update the control setting for one or more wind turbines T among the wind turbines T to be optimized based on the evaluation value L (step S104). The second output optimization program 17 updates the control settings for one or more wind turbines T to be optimized based on the evaluation value L by causing the wind farm controller 10 to perform the above processing. Functions as a setting update unit.
In some embodiments, the wind farm controller 10 obtains each output of the leeward wind turbine T after the delay time t has elapsed from the time when the control setting of the setting change wind turbine T is changed in the step of obtaining the evaluation value L. The total output of the wind turbine T to be optimized may be calculated and the evaluation value L may be obtained from the calculation result of the total output.

  According to said structure, when the control setting of the setting change windmill T is changed among the optimization target windmills T, it sets about each of the one or more leeward side windmills T located in the downstream of the setting change windmill T. A delay time t until the influence of the setting change applied to the changed wind turbine T reaches each leeward wind turbine T is calculated. Further, the total output of the wind turbine T to be optimized is calculated using the outputs of the leeward wind turbines T after the delay time t has elapsed since the control setting change time of the setting change wind turbine T, and the total output is calculated. An evaluation value L is obtained based on the result. Based on the obtained evaluation value L, the control setting is updated for one or more wind turbines T among the wind turbines T to be optimized. Therefore, the influence of the setting change applied to the upstream wind turbine T is determined based on the evaluation value L obtained in consideration of the output of the leeward wind turbine T at the timing of reaching the downstream wind turbine T. Since the control setting is changed, a more appropriate evaluation value L in accordance with the actual situation of the wind farm 1 can be obtained. Accordingly, it is possible to set an optimization target for the total power output of the entire wind farm 1 while appropriately taking into consideration the influence on the downstream wind turbine T due to the setting change applied to the upstream wind turbine T. Further, by additionally applying such a wind farm controller 10 to an existing wind farm, the output optimization of the wind farm 1 can be effectively realized with a small-scale modification.

  As shown in FIG. 13, in some embodiments, at least one setting change windmill T is located downstream of the first setting change windmill T (for example, windmill T1) and the first setting change windmill T. A second setting change windmill T (for example, windmill T2) may be included. In some embodiments, the wind farm controller 10 changes the control setting (step S101) until the influence of the change in the control setting of the first setting change windmill T reaches the second setting change windmill T. The previous value of the control setting of the second setting change windmill T is held (step S111), and when the influence of the change of the control setting of the first setting change windmill T reaches the second setting change windmill T (for example, at time t = T1 + D1 / V2), the control setting of the second setting change wind turbine T may be changed (step S112).

  According to said structure, until the influence of the change of the control setting in the 1st setting change windmill T reaches | attains the 2nd setting change windmill T located in the downstream of a wind direction rather than this 1st setting change windmill T, In the 2 setting change wind turbine T, the previous control set value is held. That is, the control setting of the second setting change wind turbine T is changed before the delay time t until the influence of the control setting change in the first setting change wind turbine T reaches the second setting change wind turbine T elapses. The control setting of the second setting change wind turbine T is changed after the delay time t has elapsed. Therefore, with respect to the influence of the control setting change on the first setting change windmill T, the control setting can be changed at an appropriate timing in the second setting change windmill T. Thereby, optimization of the power generation total output of the whole wind farm 1 can be aimed at.

  In some embodiments, the delay time t is calculated based on the distance D between the setting change wind turbine T and the downstream wind turbine T and the wind speed V (wake wind speed) downstream of the setting change wind turbine T. May be. In this way, it is possible to make the accuracy when calculating the delay time t until the influence of the change of the control setting of the setting change wind turbine T reaching the downstream wind turbine T becomes more appropriate. Thereby, optimization of the power generation total output of the whole wind farm 1 can be aimed at.

  In some embodiments, the control setting may include at least one of a power generation output command, a pitch angle command, or an orientation command of each windmill T. In this way, the delay time t until the influence of the change in the control setting including at least one of the power generation output command, the pitch angle command, or the direction command of each wind turbine T reaches the downstream wind turbine T is appropriately considered. However, the total power generation output of the entire wind farm 1 can be optimized.

  According to some embodiment mentioned above, the wind farm 1 provided with the several windmill T and the wind farm controller 10 which controls these several windmills T is implement | achieved. That is, it is possible to set an optimization target for the total power output of the entire wind farm 1 while appropriately taking into account the delay time t of the influence on the downstream wind turbine T due to the setting change applied to the upstream wind turbine T. .

In some embodiments, the wind farm controller 10 takes into account the separation distance D and the wind speed V between the wind turbines T when optimizing the wind power generator according to the optimization target. Evaluation of the optimization target (evaluation function value L) may be started after the influence of the change has arrived.
Specifically, it is assumed that the wind speed flowing into the windmill T1 at time t = t1 is V1, and optimal control setting is performed on the windmill T1 at time t = t1. Since there is a time delay for the influence of the wind turbine T1 changing the control setting to be transmitted after the wind turbine T2, it is not desirable to change the control setting of the wind turbine T2 to the newly determined optimal control setting for a certain period of time. For this reason, it is desirable that the calculation of the optimization target (value of the evaluation function L) is also started after a certain period of time.

In some embodiments, when the wind speed of the wind turbine T1 after changing the control setting of the wind turbine T1 to the optimized control setting is V2, the optimization target (evaluation function value) calculation start of the wind turbine T2 is It is desirable that time t = t1 + D1 / V2. Similarly, the optimization target (evaluation function value) calculation start of the wind turbine T3 is desirably time t = t1 + D1 / V2 + D2 / V3, and the same applies to the wind turbines T4 to Tn thereafter. Further, since V2, V3,... Change from moment to moment, it is more desirable to consider this change.
In some embodiments, as an optimization target (value of the evaluation function L), an average total output obtained by summing time averages of the power generation outputs (P1, P2, P3,...) Of each wind turbine may be used. Good.
With the above configuration, the optimization target (value of the evaluation function L) can be calculated by considering the transmission delay (delay time t) until the influence of the control setting change of the upstream wind turbine T is transmitted to the downstream wind turbine T. Qualified. Thereby, the control setting of each windmill T is optimized, and the total power generation output of the wind farm 1 can be improved.

In some embodiments, the local controller LC performs control to prevent at least one of overspeed and overpower of each windmill T or at least one control of cutout control of each windmill T. It may be configured.
Specifically, in some embodiments, the local control device LC reads the safety control program 55 stored in the ROM 53, expands it in the RAM 52, and executes it, for example, as shown in FIG. The control for preventing at least one of the overspeed and the overpower of each windmill T, or at least one control of the cutout control for each windmill T may be performed (step S201). The safety control program 55 may function as a safety control unit by causing the wind farm controller 10 to realize the above processing.

In some embodiments, the wind farm controller 10 performs an optimization process for a plurality of wind turbines T, and based on a result of the optimization process, a limit value or direction regarding each pitch angle or output of each wind turbine T. You may comprise so that at least one of the command value regarding an angle | corner or a wind direction deviation may be given to each local control device LC.
Specifically, in some embodiments, the wind farm controller 10 may perform optimization processing for a plurality of wind turbines T (step S202). In some embodiments, the wind farm controller 10 sets at least one limit value of the pitch angle, output, azimuth angle, or wind direction deviation of each windmill T based on the result of the optimization process to each local controller LC. (Step S203).
In some embodiments, the local controller LC relates to a limit value related to the pitch angle or output of each windmill T, or to an azimuth angle or wind direction deviation based on the limit value given from the wind farm controller 10 in step S203. A process of updating at least one of the command values may be executed (step S204).

  If comprised as mentioned above, the control which prevents at least one of the overspeed or overpower of each windmill T by the some local control apparatus LC each provided corresponding to the some windmill T, or each windmill T At least one of the cutout controls is performed. In addition, the wind farm controller 10 performs optimization processing for a plurality of wind turbines T, and based on the results of the optimization processing, commands relating to the pitch angle or output limit value, azimuth angle, or wind direction deviation of each wind turbine T. At least one of the values is given to each local controller LC. In other words, the wind farm controller 10 gives one or more of the limit value related to the pitch angle or output of each wind turbine T or the command value related to the azimuth angle or wind direction deviation to each local control device LC. Thereby, it is possible to additionally install the wind farm controller 10 that optimizes the power generation output without changing the function of each local control device LC with respect to the existing wind farm. Therefore, the total power output of the wind farm 1 can be improved through the optimization process of each windmill T without affecting the safety of each windmill T in the existing wind farm.

  In some embodiments, each local controller LC may be configured to perform primary processing of the operating data of each windmill T and send the results of the primary processing to the wind farm controller 10. Specifically, in some embodiments, each local control device LC performs a primary process of operation data of each windmill T (step S205), for example, as shown in FIG. The result may be sent to the wind farm controller 10 (step S206). In this way, the primary processing of the operation data of each windmill T is performed by each local control device LC, and the result of the primary processing is sent to the wind farm controller 10. Therefore, it is not necessary to perform primary processing of the operation data of each windmill T in the wind farm controller 10 connected to the plurality of local control devices LC. Thereby, since the calculation load in the wind farm controller 10 can be reduced, an optimization process can be performed smoothly.

In some embodiments, the wind farm controller 10 starts data processing to send a data processing start command to a plurality of local controllers LC so as to start primary processing of operation data including outputs of a plurality of wind turbines T. A command unit (not shown) may be included, and each local control device LC performs primary processing of the operation data of each wind turbine T based on the data processing start command transmitted from the data processing start command unit, and performs primary processing. May be configured to be sent to the wind farm controller 10.
Specifically, in some embodiments, the wind farm controller 10 includes a plurality of local controls to start primary processing of operation data including outputs of a plurality of windmills T, for example, as shown in FIG. A data processing start command may be sent to the device LC (step S211). Each local control device LC performs primary processing of the operation data of each wind turbine T in response to the data processing start command transmitted from the wind farm controller 10 (step S212), and the result of the primary processing is displayed on the wind farm. -You may send to the controller 10 (step S213). If it does in this way, according to the data processing start instruction | command transmitted by the wind farm controller 10 as a data processing start instruction | indication part, each local control apparatus LC will perform the primary process of the operation data containing the output of several windmills T The result of the primary processing is transmitted from each local control device LC to the wind farm controller 10. Thereby, the operator of the wind farm controller 10 can collect the operation data of each windmill T at a necessary timing as necessary, and, for example, based on the collected result of the primary processing, At least one of the limit value regarding the pitch angle or the output, or the command value regarding the azimuth angle or the wind direction deviation can be transmitted to each local control device LC.

In some embodiments, the wind farm controller 10 performs the optimization process (see step S202), for example, as shown in FIG. 12 described above, at least one of the wind turbines T to be optimized. The control setting of the setting change wind turbine T is changed (see step S101), and the setting change is made for each of the one or more leeward wind turbines T located downstream of the setting change wind turbine T among the wind turbines T to be optimized. A delay time t until the influence of the control setting change of the wind turbine T reaches the leeward wind turbine is calculated (see step S102), and an evaluation value L including at least the total output of the wind turbine T to be optimized is obtained (step S103). Reference), based on the evaluation value L, the control setting may be updated for one or more wind turbines T of the optimization target wind turbines T (see step S104).
In some embodiments, the wind farm controller 10 is delayed from the change point of the control setting of the setting change wind turbine T when obtaining the evaluation value L (see step S103), for example, as shown in FIG. Even if the total output of the wind turbine T to be optimized is calculated using the outputs of the leeward wind turbines T after the time t has elapsed (see step S111), the evaluation value L can be obtained from the calculation result of the total output. Good (see step S112).

  According to some embodiments described above, the wind farm controller 10 has a plurality of network devices via the network hub 3 as a control device different from the SCADA server 4 that collects information from the plurality of local control devices LC. Each is connected to a local control device LC. That is, by providing the wind farm controller 10 separately from the SCADA server 4, a plurality of local control devices LC and SCADA servers 4 respectively corresponding to the plurality of wind turbines T in the existing wind farm are relayed intensively. The operation of the existing wind farm can be optimized with a simple configuration in which the wind farm controller 10 is additionally installed in the network hub 3 without changing the function of the network hub 3 to be performed.

Note that the wind farm controller 10 according to some embodiments described above may include a plurality of information processing apparatuses. These information processing apparatuses may perform these processes in a distributed manner.
Further, by recording a program for executing each process of the above-described embodiments on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium, You may perform the various process mentioned above.

  Here, the “computer system” may include hardware such as an OS (Operating System) and peripheral devices. Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW (World Wide Web) system is used. The “computer-readable recording medium” refers to a portable disk such as a flexible disk, a magneto-optical disk, a ROM (Read-only Memory), a writable nonvolatile memory such as a flash memory, a CD (Compact Disc) -ROM, or the like. A storage device such as a medium or a hard disk built in a computer system.

  Further, the “computer-readable recording medium” refers to a volatile memory (for example, DRAM (Dynamic) in a computer system serving as a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. Random Access Memory)) that holds a program for a certain period of time is also included. The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  The present invention is not limited to the above-described embodiments, and includes forms obtained by changing the above-described embodiments and forms obtained by combining these forms.

1 Wind Farm 2 Signal Line 3 Network Hub 4 SCADA Server (Central Control Unit)
10 Wind farm controller (wind farm controller)
11 CPU
12 RAM
13 ROM
14 bus 15 output calculation program 16 first output optimization program (operation control program / perturbation applying unit / evaluation value calculation unit / update unit)
17 Second output optimization program (operation control program / setting change unit / delay time calculation unit / evaluation value calculation unit / setting update unit)
18 Database 21 Yaw Motor 22 Yaw Brake Drive Actuator 23 Pitch Drive Actuator 24 Pitch Brake Drive Actuator 31 Wind Direction Sensor 32 Wind Speed Sensor 33 Load Sensor 40 Controller 41 Plant (Control Object)
42 Control target 43 Optimization tool 51 CPU
52 RAM
53 ROM
54 Bus 55 Safety control program (safety control section)
102 Blade 104 Hub 105 Rotor (Rotating blade)
106 Main shaft 107 Main shaft bearing 108 Drive train 109 Output shaft 110 Generator 112 Nacelle 114 Nacelle base plate 116 Tower 118 Yaw slewing bearing 119 Yaw slewing mechanism T, T1 to Tn Windmills LC1 to LCn Local controller

Claims (11)

Multiple windmills,
A plurality of wind turbines each provided for each of the plurality of wind turbines to prevent at least one of overspeed or over power of each of the wind turbines, or to perform at least one control of cutout control of each of the wind turbines. A local control unit;
Performing optimization processing of the plurality of wind turbines, and at least one of a limit value related to the pitch angle or output of each wind turbine or a command value related to azimuth angle or wind direction deviation based on the result of the optimization processing A wind farm controller configured to provide to the local controller of
A wind farm comprising: A central control unit for collecting information from at least the plurality of local control devices;
A network hub provided between the central control unit and the plurality of local control devices;
Further comprising
The wind farm control device according to claim 1, wherein the wind farm control device is connected to the plurality of local control devices via the network hub.   3. Each of the local control devices is configured to perform primary processing of operation data of each of the wind turbines and send a result of the primary processing to the wind farm control device. The listed wind farm. The wind farm control device
A setting change unit for changing a control setting of at least one setting change wind turbine among the wind turbines to be optimized;
The influence of the change in the control setting of the setting change wind turbine reaches the leeward wind turbine for each of one or more leeward wind turbines located downstream of the setting change wind turbine among the wind turbines to be optimized. A delay time calculation unit for calculating the delay time until
An evaluation value calculation unit for obtaining an evaluation value including at least the total output of the wind turbine to be optimized;
A setting update unit for updating control settings for one or more wind turbines among the wind turbines to be optimized based on the evaluation value;
Including
The evaluation value calculation unit calculates the total output of the wind turbine to be optimized using the output of each of the leeward wind turbines after the delay time has elapsed since the change of the control setting of the setting change wind turbine. The wind farm according to any one of claims 1 to 3, wherein the evaluation value is obtained from a calculation result of the total output. The wind farm control device includes a data processing start command unit for sending a data processing start command to the plurality of local control devices so as to start primary processing of operation data including outputs of the plurality of wind turbines,
5. Each of the local control devices is configured to perform the primary processing of the operation data of each of the wind turbines and send a result of the primary processing to the wind farm control device. The listed wind farm. The at least one setting change windmill is:
A first setting change windmill;
A second setting change wind turbine located downstream of the first setting change wind turbine;
Including
The setting change unit
Until the influence of the change of the control setting of the first setting change wind turbine reaches the second setting change wind turbine, the previous value of the control setting of the second setting change wind turbine is held,
The control setting of the second setting change wind turbine is changed when an influence of the change of the control setting of the first setting change wind turbine reaches the second setting change wind turbine. Item 6. The wind farm according to Item 4 or 5.   The wind farm according to any one of claims 4 to 6, wherein the control setting includes at least one of a power generation output command, a pitch angle command, and a direction command of each of the wind turbines. A wind farm operating method comprising: a plurality of wind turbines; a plurality of local control devices provided for the plurality of wind turbines; respectively; and a wind farm control device for performing optimization processing of the plurality of wind turbines. ,
Performing at least one of control of preventing at least one of overspeed or overpower of each of the wind turbines, or cutout control of each of the wind turbines, by each of the local control devices;
Performing the optimization process of the plurality of wind turbines by the wind farm control device;
Based on the result of the optimization process, at least one of a limit value related to the pitch angle or output of each windmill or a command value related to azimuth angle or wind direction deviation is sent from the wind farm control device to each local control device. Giving step,
A method for operating a wind farm, comprising: Performing a primary process of operation data of each of the wind turbines by each of the local control devices;
The method of operating a wind farm according to claim 8, further comprising a step of sending a result of the primary processing to the wind farm control device. In the step of performing the optimization process,
Change the control setting of at least one setting change wind turbine among the wind turbines to be optimized,
The influence of the change in the control setting of the setting change wind turbine reaches the leeward wind turbine for each of one or more leeward wind turbines located downstream of the setting change wind turbine among the wind turbines to be optimized. Calculate the delay time until
Obtain an evaluation value including at least the total output of the wind turbine to be optimized,
Based on the evaluation value, while updating control settings for one or more windmills among the windmills to be optimized,
When obtaining the evaluation value, the total output of the wind turbine to be optimized is calculated using the output of each of the leeward wind turbines after the delay time has elapsed since the change of the control setting of the setting change wind turbine. The wind farm operating method according to claim 8 or 9, wherein the evaluation value is obtained from a calculation result of the total output. The wind farm control device sends a data processing start command to the plurality of local control devices to start primary processing of operation data including outputs of the plurality of wind turbines,
11. The wind farm according to claim 10, wherein each of the local control devices performs the primary processing of the operation data of each of the wind turbines, and sends a result of the primary processing to the wind farm control device. how to drive.