KR20190087095A - Method and system for yaw control of WindTurbin - Google Patents

Abstract

The present invention relates to a method and a system for yaw control of a wind turbine which can minimize a control parameter of a yaw control system (YCS) to simultaneously minimize a yaw error and use of a yaw actuator. The method for yaw control of a wind turbine comprises: a step of constructing a first yaw control system for performing a yaw control operation based on a normal wind direction measurement result and a second yaw control system for performing a yaw control operation based on an indirect wind direction measurement result; a step of defining an objective function and constraints for yaw error and yaw actuator minimization for the first and the second yaw control system; a step of optimizing each parameter of the first and the second yaw control system by data mining based on multi-objective particle swarm optimization (MOPSO) to acquire an optimal solution for each yaw control system; a step of comparing and analyzing the optimal solution for each yaw control system to generate a performance comparison analysis table; and a step of referring to the performance comparison analysis table if a system driving condition is inputted to select and drive a yaw control system satisfying the system driving condition.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wind turbine,

The present invention relates to a yaw control method and system for a wind turbine, and more particularly, to a yaw control method and system for a wind turbine that can optimize control parameters of a yaw control system so as to simultaneously minimize the use of yaw errors and yaw actuators .

Recently, interest in wind turbine (WT), which is the main equipment to acquire and develop wind power, has been focused on optimizing performance of wind turbine. In order to ensure high performance while minimizing costs, an optimal solution for wind turbines is continuously being developed. Among various solutions, control technology plays an essential role in directly affecting the power production and component load of the wind turbine.

Wind turbines generally have three control actuators, such as pitch actuators, torque actuators and yaw actuators. Pitch actuators and torque actuators are regarded as two dominant actuators because they can provide fast response to wind speed changes, and many technical developments have been made, but the development of technology related to yaw control systems is relatively sluggish .

The yaw control system (YCS) aims at minimizing the yaw error, which is the difference between the wind direction and the nacelle position, and minimizing the use of the yaw actuator. The increase in yaw rate reduces the wind force extracted by the horizontal axis WT, affects the structural load of the component, and the operation of the yaw actuator directly affects the life of the yaw motor and yaw brakes and consumes power .

However, since minimizing the yaw error requires continuous operation of the yaw actuator, minimizing the yaw error is inconsistent with the reduction in yaw actuator usage.

In the case of industrial WT, there are limitations that can not sufficiently meet the two goals of YCS. According to studies on WT operation, the static urine error is about 10 degrees and 5 degrees, respectively, when the wind speed is less than 20m / s, and the failure rate accounts for about 12.5% of the total failure rate. Therefore, there is a need to improve the performance of YCS. Because the YCS's hardware is pre-engineered using advanced manufacturing techniques, performance is entirely dependent on the control strategy employed.

The control strategy of YCS based on wind direction measurement technology can be classified into 4 types such as 1) control without wind direction measurement, 2) control of normal measurement, 3) control of advanced measurement, and 4) control through indirect measurement.

Of these four types, the second type using wind-induced measurement information is widely used in industrial WTs, but performance is suffering due to inaccurate measurements interrupted by WT operation. Alternatively, the first type of direct retrieval of the maximum power point (MPP) has been proposed, but since the MPP of the WT depends on wind speed and direction, applying this type of control makes it very difficult There is a big limit. Recently, a third type of obtaining wind direction information using Lidar and ultrasonic (Sodar) technology has recently been proposed, but it requires a costly measuring device and is limited to high power WT only. Compared with the third type, the fourth type of control using predicted future information provided by some prediction algorithms provides similar performance and is cost effective. With this in mind, we show that there are two favorable types of control strategies for YCS, namely normal and indirect measurements.

The control strategy through normal measurement is relatively simple, and the yaw actuators are activated when the yaw error measured by the weather vane exceeds some threshold. Nevertheless, the measured wind direction always includes high frequency noise and singularity. To solve this problem, an averaging or low-pass filter is used to provide a reference to the yawing motion. On the other hand, to avoid excessive use of yaw actuators, yaw systems are always activated at discrete intervals using predefined logic controls. A typical control strategy is used by the NREL CART3 (Controls Advanced Research Turbine 3-Bladed) turbine [prior art document 1: Kragh KA; Fleming PA. Rotor Speed Dependent Yaw Control of Wind Turbines Based on Empirical Data. AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition. 2012]. Several parameters are used in the control strategy but are predefined in an unoptimized manner. Investigation of parameter optimization is important to understand the potential performance of this control strategy.

On the other hand, the indirect measurement control is more flexible and allows not only simple control but also complex control. A predictive model of the YCS can be constructed using short-term predictions of the wind direction provided by advanced forecasting techniques. Therefore, sophisticated control based on predictive models is easy to adopt. Among them, the model predictive control (MPC) is YCS [Prior Art Document 2: Song D, Yang J, Fan X, Liu Y, Liu A, Chen G, and Joo YH. Maximum power extraction for wind turbines through a novel yaw control solution using predicted wind directions. Energy Conversion and Management 2018; 157: 587-99.

Because MPC can predict future behavior using future plant models, it is considered a natural choice for WT's control strategy and can solve optimal control problems while respecting the constraints imposed by WT's designers.

On the other hand, parameter determination is an important part of the MPC design, but there is a publication that states that the MPC can be helpful to the YCS, but it does not support the parameter selection procedure.

1. Kragh KA; Fleming PA. Rotor Speed Dependent Yaw Control of Wind Turbines Based on Empirical Data. AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition. 2012. 2. Song D, Yang J, Fan X, Liu Y, Liu A, Chen G, Joo YH. Maximum power extraction for wind turbines through a novel yaw control solution using predicted wind directions. Energy Conversion and Management 2018; 157: 587-99. 3. Coello CAC, Lechuga MS. MOPSO: A Proposal for Multiple Objective Particle Swarm Optimization. In Proceedings of Congress on Evolutionary Computation. 2002; 2: 1051-56.

In order to solve the above problems, the present invention is to provide a yaw control method and system for a wind turbine that can optimize control parameters of a YCS to simultaneously minimize yaw error and yaw actuator use.

Another object of the present invention is to provide a yaw control method and system for a wind turbine that has a plurality of YCSs and can selectively drive only YCSs suited to system drive conditions by grasping the performance of each of a plurality of YCSs.

In addition, the present invention is intended to provide a yaw control method and system for a wind turbine that can detect the influence of wind direction changes.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

As a means for solving the above-mentioned problems, according to an embodiment of the present invention, there is provided a method for controlling a yaw control system, comprising: a first yaw control system for performing a yaw control operation based on a wind direction normal measurement result; Configuring a yaw control system; Defining an objective function and constraints for minimizing a yaw error and a yaw actuator for each of the first and second yaw control systems; Acquiring an optimal solution for each of the first and second yaw control systems by performing parameter optimization of each of the first and second yaw control systems through data mining based on Multi-Objective Particle Swarm Optimization (MOPSO); Comparing and analyzing the optimum solution for each yaw control system to generate a performance comparison analysis table; And selecting and operating one yaw control system satisfying the system driving condition with reference to the performance comparison analysis table when a system driving condition is input.

Wherein the first yaw control system adjusts a nacelle position and compares the nacelle position with a wind direction to calculate a yaw error; And low pass filtering the urine error to a first time T _fast and a second time T _slow and T _slow > T _fast to obtain a first change measurement value er _fast and a second change measurement value er _slow , (N) of the nose system is shifted based on the second measured value er _slow when the integration result of the first change measurement value er _fast reaches the threshold value Th Controller.

Wherein the second yaw control system adjusts the nacelle position and compares the nacelle position with the wind direction to calculate a yaw error; An ideal prediction module for calculating a wind direction prediction value based on the nacelle position and the yaw error, a prediction model for calculating a yaw error prediction value based on the wind direction prediction value and the nacelle position, an objective function for minimizing the use of yaw error and yaw actuator And a second yaw controller including an objective function minimizing unit for predicting and providing a nacelle position value in which the yaw error prediction value is minimized through the objective function minimization.

Wherein the yaw control unit includes: an ideal prediction module for updating a wind direction prediction result of each sampling period to calculate a wind direction prediction value at a next sampling time; A prediction module for calculating a yaw error prediction value at a next sampling time based on the wind direction prediction value, the nacelle position, and a nacelle allowable speed value at a next sampling time; And a quality function for predicting and providing a nacelle position value in which the yaw error prediction value is minimized through minimizing the first and second objective functions after the first and second objective functions are calculated to calculate the yaw error and the yaw actuator use amount, And a minimization unit.

Wherein performing the parameter optimization comprises: generating an initial model including wind direction data, a yaw control system, a MOPSO parameter, and an objective function; Repeatedly performing a simulation test based on the wind direction data, the yaw control system, and the MOPSO parameter using an initial control parameter of the MOPSO algorithm; And sequentially evaluating each entity of the population based on the objective function and the constraint, and then acquiring and storing the non-dominant solution as an optimal solution.

"으로 표현되고, 상기 제2 목적 함수는 ＂

＂로 표현되며, 상기 제한 조건은 '

" 및 "

"인 것을 특징으로 한다. Wherein when the yaw control system is the first yaw control system, the objective function is constituted by a first objective function for minimizing a yaw error and a second objective function for minimizing a yaw actuator usage amount,

Quot; and the second objective function is expressed as "

Quot ;, and the constraint is expressed as "

"And"

"

"로 표현되고, 상기 제2 품질 함수는 ＂

＂로 표현되고, 상기 제한 조건은 "

" 및 "

"이며, 상기 θ_ye는 요 오차, 상기

는 요 속도, 상기 w₁ 및 w₂는 목적 함수 최적화를 통해 조정되는 가중치인 것을 특징으로 한다. Wherein when the yaw control system is the second yaw control system, the objective function is constituted by a first quality function for minimizing a yaw error and a second quality function for minimizing a yaw actuator usage amount,

Quot; and the second quality function is expressed as "

Quot; and the constraint is expressed as "

"And"

Quot ;, and &thetas; _ye denotes a yaw error,

Is a yaw rate, and w ₁ and w ₂ are weights adjusted through objective function optimization.

The method may further include the step of calculating the sensitivity according to the wind direction condition for each yaw control system by optimizing each of the first and second yaw control systems while comparing and analyzing the optimization results while varying the wind direction conditions of the wind direction data .

The method may further include re-executing only the parameter optimization of the first yaw control system when the wind direction condition of the wind direction data is changed.

According to another aspect of the present invention, there is provided a first yaw control system for performing a yaw control operation based on a wind direction normal measurement result; A second yaw control system for performing a yaw control operation based on the wind direction indirect measurement result; The objective function and the restriction for minimizing the yaw error and the yaw actuator are defined for each of the first and second yaw control systems, and then the first and second yaw control motions are performed through data mining based on Multi-Objective Particle Swarm Optimization (MOPSO) An optimization performing unit for acquiring an optimum solution for each yaw control system by performing parameter optimization of each control system; A performance comparison and analysis unit for comparing and analyzing the optimal solution for each yaw control system to generate a performance comparison analysis table; And a system selecting unit for selecting and operating one yaw control system satisfying the system driving condition by referring to the performance comparison analysis table when the system driving conditions are inputted.

The present invention proposes a new MOPSO-based data mining algorithm for optimizing control parameters of YCS. Considering the fact that a wind farm can have certain wind conditions, we collect the wind direction history data and use it as an input to the MOPSO algorithm to obtain the optimization parameters of the two YCSs. With MOPSO, you can easily discard invalid solutions, so you can efficiently search for optimal solutions.

The present invention also examines the potential performance of both YCSs. According to the existing invention, MPC using normal wind direction in two aspects of reducing the yaw error and actuator usage may be more important than logic control using normal measurement, but the ultimate potential of improved performance as a result of parameter optimization is compared and quantified It does not. By setting parameter limits to performance limits for various control strategies, you can identify which control strategies will potentially provide more benefits to the yaw control system.

In addition, the present invention investigates the effect of wind direction conditions on parameter optimization and potential performance of both YCSs. Since the YCS aims to track the wind direction, wind direction conditions can have a significant impact on YCS. However, in the past, this important problem has not been considered at all. Accordingly, the present invention enables to understand and utilize the influence of wind direction conditions on YCS.

1 is a view for explaining a yaw control method of a wind turbine according to an embodiment of the present invention.
2 is a diagram illustrating a configuration of a first YCS according to an embodiment of the present invention.
3 is a diagram illustrating a configuration of a second YCS according to an embodiment of the present invention.
4 is a diagram for explaining a parameter optimization method for searching for optimal parameters through repeated simulation using wind direction history data as an input.
5 is a diagram for explaining the sensitivity of YCS according to a change in wind direction condition.
6 is a diagram illustrating a yaw control integrated system of a wind turbine according to an embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features and advantages of the present invention will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

The following terms are defined in consideration of the functions of the present invention, and these may be changed according to the intention of the user, the operator, or the like.

The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. These embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art to which the present invention pertains. Only. Therefore, the definition should be based on the contents throughout this specification.

1 is a view for explaining a yaw control method of a wind turbine according to an embodiment of the present invention.

Referring to FIG. 1, the method of the present invention includes a first yaw control system that performs a yaw control operation based on a wind direction normal measurement result, and a second yaw control system that performs a yaw control operation based on a wind direction indirect measurement result (S2) of defining an objective function and a restriction for minimizing a yaw error and a yaw actuator, for each of the first and second yaw control systems, a step (S2) for generating a multi-objective particle swarm optimization (MOPSO) (S3) of acquiring an optimum solution for each of the first and second yaw control systems by performing parameter optimization of each of the first and second yaw control systems through mining, comparing and analyzing the optimal solutions for each of the yaw control systems, (S4), and when a system driving condition is inputted (S5), a requirement It may include a step (S6), such as to select and drive the system.

Hereinafter, with reference to Figs. 2 to 5, the method of the present invention will be described in more detail.

First, step S1 will be described in more detail as follows.

The yaw system can maximize wind power by aligning the blade rotor with the wind direction. However, the yawing system is implemented as a low-speed, bulky machinery, and the speed of wind direction tracking is inevitable. In order to improve wind direction tracking capability, a yaw control system (YCS) in which a control strategy plays a key role should be appropriately designed.

Accordingly, the present invention proposes two types of YCS, and selectively drives the two types of YCS according to the system driving conditions, thereby ensuring the highest performance more stably.

The performance of the YCS can be evaluated through the following indices.

YCS wind-force extraction and yaw actuator use, which can be evaluated by examining the relative data of yaw error and yaw operation time, respectively. Yaw error (θ _ye) is the difference of the wind direction (θ _wd) and nacelle position (θ _np), can be formulated as equation (1).

[Equation 1]

The yaw error can be used to evaluate wind direction tracking performance, but its statistical value (eg, mean absolute error, mean square error) is more efficient to evaluate overall performance over a specific time period. However, such a yaw error value can not express the power extraction capacity of YCS explicitly. The yaw error (? _Ye ) and _Pout (power output) have a non-linear relationship and can be expressed as Equation (2).

&Quot; (2) "

In this case, ρ is the air density, A _r is the rotor area, C _p is the aerodynamic power coefficient, V ₀ is the free stream wind speed, and k is the estimated constant value, For example 1.8.

The power reduction factor ([zeta]) to better evaluate the effect of the yaw error ([theta] _ye ) on P _out (power output) under different YCSs is calculated based on Equation (3).

&Quot; (3) "

At this time, T _e is the evaluation time length.

는 요 속도, 요 동작 시간은 요 동작시에 증가된다. The yaw operation time (t _yaw ) is expressed by Equation (4)

The yaw rate and yaw operation time are increased during yaw operation.

&Quot; (4) "

In the case of a typical wind turbine, the operating mechanism of the YCS can be described as follows.

Step 1: The weather vane and nacelle position transducer measure the yaw error and nacelle position, respectively, and send it to the input to the yaw control.

Step 2: The yaw controller processes the input and determines the output of the relay used to turn the yaw motor on and off.

Step 3: The nacelle is moved by a motor driven gear.

Observed in the above step, it can be seen that the performance is determined by the adopted control strategy. As described above, two prior types of control strategies have been proposed in the prior art.

Therefore, in the present invention, a first YCS that performs a yaw control operation based on the wind direction normal measurement result based on the prior art documents 1 and 2, and a second YCS that performs a yaw control operation based on the wind direction indirect measurement result, .

2 is a diagram illustrating a configuration of a first YCS according to an embodiment of the present invention.

As shown in FIG. 2, the first YCS 100 of the present invention includes a yaw system 110 and a yaw controller 120. The yaw system 110 adjusts the position of the nacelle and calculates and reports the yaw error by measuring and comparing the position of the nacelle with the wind direction and the yaw controller 120 calculates the yaw rate at the first time T _fast and the second time the integration of the T _slow, T _slow> T _fast) with low-pass filtering by the first change measurements (er _fast) and second change measurements (er _slow) the formation, and the first change in measured values (er _fast) And moves the nacelle position of the urine system based on the second measured value er _slow when the threshold value Th is reached.

That is, the yaw error (θ _ye) two low-pass filtered by the time constant T _fast = first change measure having 1 second (er _fast) and the time constant T _slow = having 60 seconds, the second change measurements (er _slow ), and er _fast is integrated and monitored. And the first change measurement value (er _fast ) When the integration error (AccErr) reaches a threshold value (Th = 6000) that deviates by 10 degrees for 10 minutes, the nacelle position is moved to the setting position determined by er _slow .

3 is a diagram illustrating a configuration of a second YCS according to an embodiment of the present invention.

As shown in FIG. 3, the second YCS 200 of the present invention also includes a yaw system 210 and a yaw controller 220. The urine system 210 adjusts the nacelle position and compares the nacelle position with the wind direction to calculate the urine error. The yaw controller 220 includes an ideal prediction module 221 for calculating a wind direction prediction value based on a nacelle position and a yaw error, a prediction model 222 for calculating a yaw error prediction value based on the wind direction prediction value and the nacelle position, And a quality function minimizing unit 223 for predicting and providing a nacelle position value in which the yaw error prediction value is minimized by minimizing the quality function after configuring a quality function that minimizes the use of the yaw actuator.

The present invention modifies the MPC-based method including the wind direction prediction direction disclosed in the prior art document 2 in the following two aspects.

Difference 1: The present invention calculates the wind direction prediction value through the ideal prediction module instead of the ARIMA-KF module. The main purpose of the present invention is to facilitate the objective achievement by examining the potential performance of the adopted control strategy including the ideal prediction module for the control strategy. Because the prediction accuracy provided by the ARIMA-KF algorithm may be uncertain for other time series of wind direction, it is reasonable to use an ideal prediction module.

Difference 2: The quality function defined in the present invention refers to the two parts of the yaw error and the yaw time, whereas the one defined in the prior art document 2 includes the other part, that is, the yaw operation frequency. Considering that there is a linear relationship between the yaw operation time and the number of yaw operation, the number of yaw operation times is not used in the present invention.

The ideal prediction module 221 updates the wind direction prediction result of each sampling period, which is discrete data, to calculate the wind direction prediction value at the next sampling time. Only the accurate one-step result is used in the prediction model 222, not the multi-step prediction result. The output can be expressed as Equation (5).

&Quot; (5) "

는 샘플링 주기의 평균값이다. At this time,

Is the average value of the sampling period.

The prediction model 222 calculates the yaw error prediction value at the next sampling time based on the wind direction prediction value, the nacelle position, and the nacelle velocity, which can be formulated as shown in Equation (6).

&Quot; (6) "

는 다음 샘플링 시점에서의 나셀 허용 속도값이고, 수학식 7의 유한 집합으로 분류할 수 있습니다.At this time,

Is the nacelle allowable speed value at the next sampling time, and can be classified into the finite set of Equation (7).

&Quot; (7) "

The quality function minimizing unit 223 generates a quality function QF composed of first and second quality functions QF1 and QF2 for calculating a yaw error and a yaw actuator usage amount.

&Quot; (8) "

In this case, w ₁ and w ₂ are two weight values.

The minimization of QF is implemented in a "for " period, which estimates the yaw error of Equation 6 for the three admission control operations of Equation 7, evaluates it in accordance with Equation 8 and stores the minimum value QF (min) This can be explained by the following pseudo code.

Set QF (min) To inf and Set min To inf

For j = 0: 2

Calculate QF (j) Using Eqs. (6-8)

If QF (j) < QF (min) Set QF (min) To QF (j) And Set min To j

End

End

Apply

Step S2 will now be described in more detail.

In the present invention, the performance of the YCS is determined by two distinct indices, namely the minimization of the yaw error and the minimization of the actuator usage. Thus, the parameter optimization is indeed a multipurpose optimization problem (MOP), which can be formulated as in equation (9).

&Quot; (9) &quot;

In this case, X = (x ₁ , x ₂ , ..., x _n ) ^T is a decision vector belonging to a nonempty feasible region (S⊂R ⁿ ) (Denoted by f ₁ (X), f ₂ (X), ..., f _m (X)). In the MOP, F (x) is optimal if at least one of the other components can not further improve the component without degradation. In mathematical terms, f _i (X) ≤ f _i (X ^* ) for all i (i = 1,2, ..., m) and at least one j X ^* is called a non-dominated solution set or Pareto optimal if there is no other vector x∈S such that f _j (X) ≤ f _j (X ^* )). Therefore, F (X ^* ) is regarded as PF (Pareto Front). Since the MOP solution is an incomplete vector, all Pareto optimal solutions are equally desirable, with the most preferred solution being identified according to the user's preference.

At present, multi-objective optimization based on Pareto optimal theory has been intensively invented and applied to the design of intelligent evolutionary wind turbine. Among them, Partial Swarm Optimization (PSO) is recognized as a simple conceptual algorithm that is easy to implement coding and calculation efficiency.

The present invention uses the MOPSO method to find the optimal solution group according to the Pareto concept principle. Since the optimized parameters can be obtained based on the MOPSO based method, the potential performance of two YCSs can be investigated and compared based on the optimized parameters.

4 is a diagram for explaining a parameter optimization method for searching for optimal parameters through repeated simulation using wind direction history data as an input.

First, an initial model including history data of the collected wind direction, YCS, MOPSO parameters and objective function is generated (S11).

Second, the simulation test is performed using the initial control parameters provided by the MOPSO algorithm. The simulation test is continued until the wind direction history data is completely consumed (S12).

Third, each entity of the population is evaluated based on the objective function and the constraint (S13).

Finally, the optimization result is determined according to the stopping principle (that is, stopped when the maximum value of the iteration time is reached). When the stopping principle is satisfied, the non-dominant solution is stored as the ultimate optimal solution. Otherwise, the iteration continues and the evaluation result is used by the MOPSO algorithm to generate the next generation control parameter (S14).

MOPSO algorithm

The PSO algorithm was first proposed by Kennedy and Eberhart in 1995 [39, 40]. At the PSO, the algorithm initializes the solution set and updates the solution through population changes to retrieve the optimal solution. A potential solution set is a set of particles called swarms that move in search space as a collaborative operation. The velocity operator performs the movement guided by the local and social components based on Equation 10a.

Equation (10a)

Equation (10b)

At this time, t = 1, 2, ... , T _max denotes the running index of the generation, and T _max is the maximum number of the population. t is the current generation. i = 1, 2, ... , N _ip is the running index of particles of the particle, and N _ip is the particle size. D is the particle dimension, and D is the dimension size. v _id (t) represents the velocity of particle i in t-th generation. Similarly, x _id (t) represents the particle position. p _id (t) is the best position for the individual found by particle i, and p _gd (t) is the best global position ever found in the entire swarm. Constants c1 and c2 are acceleration coefficients, r1 and r2 are random values uniformly distributed in [0,1], and w is an inertia weight given in Equation 10c.

(10c)

Where w _max and w _min are the maximum and minimum inertia weights, respectively.

The main idea of the MOPSO algorithm is that the output of the PSO is a non-dominant solution or a solution set named PF. In the present invention, power reduction and yaw actuator use are simultaneously minimized by using NSPSO (Non-dominated Sorting PSO) algorithm proposed by Li in [41]. A key feature of this method is that it provides an appropriate selection pressure to propel the swarm population toward the PF, since it updates each particle by comparing its individual offspring with the individual best of all particles on the whole particle. In addition, to maintain a variety of non-dominant solution sets, the present invention uses a dense distance method of selecting Pgd (t).

Objective functions and designed control parameters

First, the first YCS 100 has the following first and second objective functions.

The parameter optimization of the YCS is aimed at minimizing the error and minimizing the actuator usage, so use the performance index. Therefore, the first objective function is expressed as Equation (11), and the amount of power reduction due to the minimization of the error is determined.

&Quot; (11) &quot;

는 s번째 샘플링 시점에서의 요 오차이다. Here, N is a sampling number of the wind direction history data,

Is the yaw error at the s-th sampling point.

The second objective function calculates the yaw actuator use amount, which is calculated through the yaw operation time expressed by Equation (12).

&Quot; (12) &quot;

Choosing designed control parameters is important for performance optimization. Control system 1 defines three control parameters (including T _fast , T _slow, and Th) that are selected by the designed control parameters, T _slow and Th, because T _fast and Th are logically connected. Considering the physical meaning, T _slow and Th are chosen in the following ranges.

&Quot; (13) &quot;

Furthermore, besides the individual ranges given in equation (13), T _fast and Th are limited to a set of achievable objective functions.

&Quot; (14) &quot;

In Equation 14, the set of control parameters is considered a dominant solution set and discarded when the power reduction percentage is greater than 10% or the yaw actuator utilization is greater than 15%.

Then, the second YCS 200 uses an MPC-based control strategy that defines two control parameters w ₁ and w ₂ , and uses the first and second quality functions described above as the first and second objective functions . These two parameters are selected as the designed control parameters, and the arrangement is preselected as follows.

&Quot; (15) &quot;

Similarly, the control parameter set is limited to Equation (16).

&Quot; (16) &quot;

Step S3 will now be described in more detail.

The wind direction data of the present invention may be actual wind direction data collected by a weather vane mounted on a nacelle of an operating wind turbine. And may be time series data repeatedly measured and collected over a predetermined period of time. For example, it may be time series data measured for 3 days and having a wind direction change value (? (? _Wd ) as shown in Table 1). At this time, the wind direction change value (? (? _Wd )) can be calculated by the following equation (17).

[Table 1]

&Quot; (17) &quot;

However, the actual wind direction data may include high frequency noise and some disturbances, but high frequency noise and disturbance may be influenced by some data processing algorithms when evaluating control performance. Therefore, the present invention allows direct use of actual wind direction data including high frequency noise and some disturbance, that is, original data, as model inputs.

The MOPSO parameter of the present invention can be selected as follows.

- Population size: N _ip = 100;

- Maximum number of generation: T _max = 100;

Acceleration coefficients: c1 = 0.8, c2 = 2.0;

- Inertia weights: w _max = 0.9, w _min = 0.4

As shown in Table 1, the wind direction change is 5.17, 4.16 and 6.54 for DAY 1,2,3 respectively. For ease of performance analysis, results based on DAY 3 wind direction data are used for detailed analysis focusing on parameter optimization and potential performance studies of two YCSs. Also, the results of DAY 1,2 are used for sensitivity analysis considering other wind direction conditions.

In this case, the first YCS 100 acquires the following optimization result value.

By performing the optimization method of FIG. 4 for the first YCS 100, initial control parameters (reference) and PF (solutions 1, 2, 3 and 4) as shown in Table 2 can be obtained.

[Table 2]

Then, by performing the optimization method of FIG. 4 for the second YCS 200 in the same manner, initial control parameters (reference) and PF (solutions 1 and 2) as shown in Table 3 can be obtained.

[Table 3]

Next, step S4 will be described in more detail as follows.

The solution obtained through the optimization of the first YCS 100 and the solution obtained through the optimization of the second YCS 200 are compared and analyzed to obtain the performance comparison analysis result as shown in Table 4 below.

From the performance comparison analysis results in Table 4, it can be clearly seen that the two control systems have different performance. It can be seen that the performance of the second YCS 200 is superior to that of the first YCS 100 when the usage amount of the actuator is higher than a specific value, and vice versa.

[Table 4]

For example, if 4.9% or more of the yaw actuator usage is allowed for the solution 2 (f ₁ = 0.0324, f ₂ = 0.049), the performance of the second YCS 200 is better, 100).

Using 14% of the actuator as in Solution 1, the power extraction efficiency of the second YCS 200 is 97.73%, which is 0.54% higher than that of the first YCS 100, and as in Solution 3, %, The power extraction efficiency of the second yaw control system is reduced to 95.45%, which is 0.92% less than that of the first YCS 100. [

Therefore, when the system driving conditions are inputted (S5), the present invention selects and drives the YCS satisfying the corresponding system driving conditions by referring to the performance comparison analysis table as described above, and thereby, the performance required by the user is stably secured (S6).

In addition, according to the present invention, the sensitivity of the yaw control system is calculated for each yaw control system by optimizing each of the first and second yaw control systems while varying the wind direction conditions of the wind direction data, ). &Lt; / RTI &gt;

5, all PF curves move from right to left as the change in direction of the wind decreases, and about 1% of the power extraction efficiency, while maintaining the same yaw actuator use, changes in the direction of wind δ (θ _wd ) To 5.17 of Day 1, and another 0.5% of the power extraction increases by δ (θ _wd ) and further decreases to 4.16 of Day 2. If the wind direction change is small, the overall performance is improved, which means that the YCS is easier to track the wind direction with less variation.

However, even when the wind direction conditions change, the results of the performance comparison of the first and second YCSs 100 and 200 are continuously maintained. For example, when the use of the yaw actuator is greater than a pre-set value, the second YCS 200 performs better than the first YCS 100, but its value decreases from 4.9% to 4.3% %.

Thus, under different wind conditions, it can be seen that the second YCS 200 can ensure higher power extraction efficiency within acceptable yaw actuator usage.

[Table 5]

This sensitivity analysis allows to investigate the effect of optimized parameters in various wind direction conditions. To this end, the optimized control parameters summarized in Tables 2 and 3 are used to obtain performance results, which are depicted in FIG. 5 and summarized in Table 5.

As a result, it can be seen that the fluctuation according to the change in the wind direction condition is reduced, and the performance of the first and second YCSs 100 and 200 is improved by parameter optimization. Nevertheless, the performance of the first YCS 100 using the designed control parameters optimized for DAY 3 deviates from the PF curves of DAY 2 and DAY 3, but not the PF curves of the second YCS 200.

From this, it can be seen that if the wind direction conditions change, the optimization parameters of the first YCS 100 should be searched, but the optimization parameters of the second YCS 200 can be used again. That is, when the wind direction condition of the wind direction data changes, it can be seen that only the parameter optimization of the first YCS 100 needs to be re-executed.

6 is a diagram illustrating a yaw control integrated system of a wind turbine according to an embodiment of the present invention.

As shown in FIG. 6, the system of the present invention includes a first yaw control system 100 for performing a yaw control operation based on a wind direction normal measurement result, a second yaw control system 100 for performing a yaw control operation based on a wind direction indirect measurement result, After defining the objective function and constraints for system 200, yaw error and yaw actuator minimization for each of the first and second yaw control systems, the first and second yaw control systems A performance comparison and analysis unit 400 for comparing and analyzing the optimal solutions for each of the yaw control systems to generate a performance comparison analysis table, When the driving condition is input, one yaw control system satisfying the system driving condition is selected and driven by referring to the performance comparison analysis table It may include system selection unit 500 and the like.

The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

Claims (10)

A first yaw control system for performing a yaw control operation based on a wind direction normal measurement result and a second yaw control system for performing a yaw control operation based on a wind direction indirect measurement result;
Defining an objective function and constraints for minimizing a yaw error and a yaw actuator for each of the first and second yaw control systems;
Acquiring an optimal solution for each of the first and second yaw control systems by performing parameter optimization of each of the first and second yaw control systems through data mining based on Multi-Objective Particle Swarm Optimization (MOPSO);
Comparing and analyzing the optimum solution for each yaw control system to generate a performance comparison analysis table; And
And selecting and operating one yaw control system satisfying the system drive condition by referring to the performance comparison analysis table when the system drive condition is inputted. 2. The system of claim 1, wherein the first yaw control system
A first urine system for adjusting a nacelle position and calculating a urine error by comparing the nacelle position with a wind direction; And
Pass filtering the urine error to a first time T _fast and a second time T _slow and T _slow > T _fast to obtain a first change measurement value er _fast and a second change measurement value er _slow And for moving the nacelle position of the urine system based on the second measured value er _slow when the integration result of the first change measurement value er _fast reaches the threshold value Th, And controlling the yaw rate of the wind turbine. 2. The system of claim 1, wherein the second yaw control system
A second urine system for adjusting a nacelle position and calculating a urine error by comparing the nacelle position with a wind direction; And
An ideal prediction module for calculating a wind direction prediction value based on the nacelle position and the yaw error, a prediction model for calculating a yaw error prediction value based on the wind direction prediction value and the nacelle position, an objective function for minimizing a yaw error and a yaw actuator use, And a second yaw control unit including an objective function minimizing unit for predicting and providing a nacelle position value in which the yaw error prediction value is minimized through the objective function minimization. 4. The system of claim 3, wherein the yaw control device
An ideal prediction module for updating a wind direction prediction result of each sampling period to calculate a wind direction prediction value at a next sampling time;
A prediction module for calculating a yaw error prediction value at a next sampling time based on the wind direction prediction value, the nacelle position, and a nacelle allowable speed value at a next sampling time; And
A first and a second objective function for calculating a yaw error and a yaw actuator usage amount and then minimizing a quality function to predict and provide a nacelle position value in which the yaw error prediction value is minimized by minimizing the first and second objective functions And a control unit for controlling the yaw control of the wind turbine. 5. The method of claim 2 or 4, wherein performing the parameter optimization comprises:
Generating an initial model including wind direction data, a yaw control system, a MOPSO parameter, and an objective function;
Repeatedly performing a simulation test based on the wind direction data, the yaw control system, and the MOPSO parameter using an initial control parameter of the MOPSO algorithm; And
Sequentially evaluating each entity of the population based on the objective function and the constraint, and acquiring and storing the non-dominant solution as an optimal solution. 6. The method of claim 5,
When the yaw control system is the first yaw control system,
Wherein the objective function is constituted by a first objective function for minimizing a yaw error and a second objective function for minimizing a yaw actuator usage amount,
The first objective function is "

Quot; and the second objective function is expressed as "

"
The restriction condition is'

"And"

&Quot;.&lt; / RTI &gt; 6. The method of claim 5,
When the yaw control system is the second yaw control system,
Wherein the objective function is constituted by a first quality function for minimizing a yaw error and a second quality function for minimizing a yaw actuator usage amount,
The first quality function is "

Quot; and the second quality function is expressed as "

"
The restriction condition is "

"And"

"
_Ye is the θ yaw error, the

Is a yaw rate, and w ₁ and w ₂ are weights adjusted through objective function optimization. The method according to claim 1,
Further comprising the step of calculating the sensitivity according to the wind direction condition for each yaw control system by optimizing each of the first and second yaw control systems while comparing the wind direction data of the wind direction data with the optimization result, A method for controlling yaw of a wind turbine. 9. The method of claim 8,
Further comprising the step of re-executing only the parameter optimization of the first yaw control system when the wind direction condition of the wind direction data is changed. A first yaw control system for performing a yaw control operation based on the wind direction normal measurement result;
A second yaw control system for performing a yaw control operation based on the wind direction indirect measurement result;
The objective function and the restriction for minimizing the yaw error and the yaw actuator are defined for each of the first and second yaw control systems, and then the first and second yaw control motions are performed through data mining based on Multi-Objective Particle Swarm Optimization (MOPSO) An optimization performing unit for acquiring an optimum solution for each yaw control system by performing parameter optimization of each control system;
A performance comparison and analysis unit for comparing and analyzing the optimal solution for each yaw control system to generate a performance comparison analysis table; And
And a system selector for selecting and operating one of the yaw control systems satisfying the system drive condition by referring to the performance comparison analysis table when the system drive condition is inputted.

Images