阅读说明：本技术 风力涡轮机的叶尖间隙、估算与控制 (Blade tip clearance, estimation and control of wind turbines ) 是由 王弼堃 王林鹏 陈林 于 2018-04-17 设计创作，主要内容包括：公开了一种叶尖间隙控制(TCC)(1000)的方法和风力涡轮机发电机(1),该风力涡轮机发电机具有转子,转子具有相对塔架(2)转动的至少一个叶片(5)、以及至少一个叶片的叶尖(7)与塔架之间的叶尖间隙(8),该方法包括：测量(1100)一组运行值(110)和一组叶片负载值(150)；根据该组运行值和该组叶片负载值估算(1200)叶尖间隙(210)；从所估算的叶尖间隙(210)生成(1300)控制指令(310)。还公开了一种计算机程序产品(400)和风力涡轮机发电机控制系统(300)。(A method of Tip Clearance Control (TCC) (1000) and a wind turbine generator (1) having a rotor with at least one blade (5) rotating relative to a tower (2) and a tip clearance (8) between a tip (7) of the at least one blade and the tower are disclosed, the method comprising: measuring (1100) a set of operational values (110) and a set of blade load values (150); estimating (1200) a tip clearance (210) based on the set of operational values and the set of blade load values; a control command (310) is generated (1300) from the estimated tip clearance (210). A computer program product (400) and a wind turbine generator control system (300) are also disclosed.)

1. A method of Tip Clearance Control (TCC) (1000) of a wind turbine generator (1), the wind turbine generator (1) having a rotor with at least one blade (5) rotating relative to a tower (2), and a tip clearance (8) between a blade tip (7) of the at least one blade (5) and the tower (2); the method comprises the acts of:

-measuring (1100) a set of operating values (110) and a set of blade load values (150) on the wind turbine generator (1);

-estimating (1200) a tip clearance (210) between the blade tip (7) and the tower (2) from the set of operating values (110) and the set of blade load values (150);

-generating (1300) control instructions (310) from the estimated tip clearance (210).

2. A method of Tip Clearance Control (TCC) (1000) in accordance with claim 1, wherein the act of estimating (1200) the tip clearance (8) comprises an act of determining:

Y_TC＝x₁+x₂·θ+x₃·ω+x₄·T+x₅·M_{blade root}+x₆·M_{SH Tilt}+x₇·F_{Push away}。

3. The Tip Clearance Control (TCC) (1000) method of claim 2, wherein tip clearance is determinedGap (1220) Y_TC(1230) Includes determining an M available through fitting_{Blade root}To determine M_{SH Tilt}E.g. two different M_{Blade root}For wind turbine generators having at least two blades, e.g. three different M_{Blade root}For wind turbine generators having at least three blades, and the like.

4. A method of Tip Clearance Control (TCC) (1000) according to claim 2 or 3, wherein the act of determining tip clearance (1220)

Y_TC＝x₁+x₂·θ+x₃·ω+x₄·T+x₅·M_{Blade root}+x₆·M_{SH Tilt}+x₇·F_{Push away}

Includes solving by using the set of measured operating values (110) and the set of measured blade load values (150):

AX＝Y

wherein

5. A method of Tip Clearance Control (TCC) (1000) according to any one of the preceding claims, wherein the act of generating (1300) control commands (310) is performed by:

-an act of generating a first command (311) based on the estimated tip clearance (210) and the set point tip clearance (215), wherein the first command (311) is a set point pitching moment command, and then

-an act of generating a second command (312) as a function of the first command (311) and the measured hub tilting moment, wherein the second command (312) is a pitch command.

6. The tip-clearance control (TCC) (1000) method according to any preceding claim, wherein the act of generating control commands (310) comprises an act of using individual pitch control with a dynamic lookup table having a lower limit and an upper limit, and wherein only the lower limit varies based on the estimated tip-clearance.

7. The method of Tip Clearance Control (TCC) (1000) according to any of the preceding claims, further comprising transforming the control commands (310) into individual blade commands.

8. A method of Tip Clearance Control (TCC) (1000) according to any one of the preceding claims, wherein the set of measured blade load values (150) is obtained from measurements of one or more blade value measuring devices (10), said one or more blade value measuring devices (10) being selected from the group consisting of:

-mechanical strain gauges

-optical fibre strain gauge

-laser device

-camera arrangement

-or a combination of the above.

9. The Tip Clearance Control (TCC) (1000) method according to claim 8, wherein the set of measured blade load values (150) is obtained in accordance with mechanical strain gauge means, wherein said mechanical means comprises Wheatstone bridge means and calibration means.

10. A computer program product (400) comprising instructions which, when executed by a computer, cause the computer to carry out the actions of one or more of claims 1-9.

11. A wind turbine generator control system (300) comprising means for performing the actions of one or more of claims 1-9.

12. A wind turbine generator (1) comprising a control system (300) according to claim 11.

13. Wind turbine generator (1) according to claim 12, characterised in that the means for measuring the set of blade load values (150) comprise strain gauge sensors (20) mounted on the inside of the surface of the blade root (6).

14. Wind turbine generator (1) according to claim 13, characterized in that the strain gauge sensors (20) are arranged in a wheatstone bridge configuration.

Technical Field

The invention relates to blade tip clearance estimation and control for wind turbine generators.

Background

Blade tip clearance is a significant challenge as wind turbine designs use longer blades in an attempt to extract more energy from the wind. Not only do extreme environmental conditions, such as wind shear, wind gusts, and turbulence, increase the likelihood of striking the tower as the blades pass through the tower, but blades are being designed to become increasingly flexible.

A Tip Clearance Control (TCC) method is introduced to activate the pitch action as the blades pass the tower. The closed-loop control method makes the system reliable and stable. The improved tip clearance not only avoids the risk in the wind field, but also benefits the blade design.

Disclosure of Invention

One object is achieved by a method, a control system for a wind turbine generator, and a computer program product for executing instructions, which will be outlined below.

One object is achieved by a method of Tip Clearance Control (TCC) for a wind turbine generator having a rotor with at least one blade rotating relative to a tower and a tip clearance between a blade tip of the at least one blade and the tower. The method may include the acts of measuring a set of operating values and a set of blade load values at a wind turbine generator.

The operational value may include a pitch angle, a generator speed, or a torque.

There may be an act of estimating a tip clearance between the blade tip and the tower based on the set of operating values and the set of blade load values.

There may be an act of generating control commands (310) from the estimated tip clearance.

The control instructions may be applied directly to the wind turbine generator.

As wind turbine rotors get larger, tip clearance issues arise and even impose limitations on the design of the wind turbine. To improve tip clearance, a Tip Clearance Control (TCC) method is designed to ensure minimum safe tip clearance. When the controller detects that the tip clearance is below a certain threshold, the TCC will provide separate pitch commands for three different pitch settings in order to bend the entire rotor plane to avoid extreme tip clearance events.

The TCC uses a tip clearance estimator as a system input. The estimator uses linear and non-linear combinations of blade load, pitch angle, generator speed, and torque to produce an estimated tip clearance.

The signals are measured from different sensors. The minimum tip clearance is as the blade passes through the tower and the algorithm or method is designed to combine and sample three blades in a window to estimate the tip clearance to produce a discrete input signal.

The TCC may have a dual PI (proportional integral) control module within the controller. The first PI controller is configured to generate how much hub pitch bending moment is required to avoid an extreme tip clearance condition. The second PI controller is configured to change the bending moment error to a respective pitch command. The expected tip clearance is translated by the first PI controller into a bending moment that will be translated by the second PI controller into a target pitch angle on the DQ axis. The pitch angle in the DQ axis is converted into three pitch commands by classical inverse DQ conversion.

The TCC function or method can be divided into three parts.

The first part may be a signal measurement used as an input to the method. The signals may include basic wind turbine performance or operational information, including pitch angle, generator speed, and torque measurements.

Different blade element loads may also be achieved. The load may be picked up by several strain gauge sensors mounted in the blade surface. The number of strain gauges will determine the accuracy of the estimated tip clearance.

The load measurement may be obtained using fiber optic sensing techniques. The load measurements may be obtained as will be outlined later.

The second part may be a tip clearance estimate.

As outlined, the TC estimation may use a linear regression based on wind turbine basic sensory information or measurements. One way to perform a linear fit is to use a least squares method.

The third part of the TCC function may be dual PI control in a closed loop method.

The TCC function or method may use two separate PI controllers in which the parameters are calibrated. The propagation parameters and the integration parameters may or will depend on the wind turbine model.

The additional part may be where the signals or commands are passed through an inverse DQ conversion to split or divide the signals or commands into three pitch signals in the case of a three-bladed rotor.

In one aspect, the tip clearance control method is an act of estimating tip clearance, including the acts of determining:

Y_TC＝x₁+x₂·θ+x₃·ω+x₄·T+x₅·M_{blade root}+x₆·M_{SH Tilt}+x₇·F_{Push away}

In one aspect, a tip clearance (1220) Y is determined_TCIncludes M available by fitting_{Blade root}(blade load moment) to determine M_{SH Tilt}(tilting moment). For a wind turbine generator with at least two blades, two different M's may be used_{Blade root}. For wind turbine generators having at least three blades, etc., three different M's may be used_{Blade root}。

In one aspect, the tip clearance is determined as

Y_TC＝x₁+x₂·θ+x₃·ω+x₄·T+x₅·M_{Blade root}+x₆·M_{SH Tilt}+x₇·F_{Push away}

Including unlocking:

AX＝Y

wherein

The equation is populated by a set of measured operating values and a set of measured blade load values.

Y_TCIs the tip clearance. I.e. the distance between the tip of the blade and the tower as the blade passes the tower. Tip clearance may be a similar measure of the distance between the tip of the blade and the tower or stationary component.

The operating values are as follows: θ is the pitch angle, ω is the generator speed, and T is the generator torque.

M_{Blade root}Is a measure of the flapping moment of the blade root. M_{SH Tilt}Is a measure of the tilting moment of the stationary hub. M_{SH Tilt}A non-linear combination of flap loads for 3 different blade roots can be obtained.

F_{Push away}Is the thrust of the rotor. F_{Push away}A linear combination of flap loads for 3 different blade roots can be obtained.

x1 through x7 are constants determined based on the operating data, for example, by regression fitting and a system of equations.

In one aspect, the act of generating a control instruction is performed by acts that will be described below.

There may be an act of generating a first command based on the estimated tip clearance and the setpoint tip clearance, wherein the first command is a setpoint pitching moment command.

This action is followed by an action of generating a second command based on the first command and the measured hub pitching moment, wherein the second command is a pitch command.

In one aspect, the act of generating the control instructions includes an act of using an IPC (individual Pitch control) with a dynamic lookup table having a lower bound and an upper bound, and wherein only the lower bound varies according to the estimated tip clearance.

The dynamic lookup table may include hub pitching moments that vary according to wind speed. An upper limit curve and a lower limit curve may be present.

The dynamics may be such that the upper limit curve and/or the lower limit curve vary in dependence of said wind speed.

In extreme wind conditions, such as wind shear and wind changing conditions, there may be advantages. In an example, if e.g. a change or a rapid change in the hub load is detected, the (individual) pitch command or set point lower limit value may be increased or the upper limit value may be decreased to predict the pitch command behavior.

In one aspect, the method further comprises translating (possibly DQ translating) the control commands (possibly pitch commands) into individual blade commands, i.e. into individual blade pitch commands.

In one aspect, a set of measured blade load values is obtained from measurements of one or more blade value measurement devices. There may be a mechanical strain gauge. There may be a fiber optic strain gauge. There may be a laser device. There may be a camera device. Combinations may be used to provide better accuracy or data redundancy.

In one aspect, a set of measured blade load values is obtained from measurements of a mechanical strain gauge apparatus, wherein the mechanical apparatus includes a Wheatstone bridge apparatus and a calibration apparatus.

The outlined acts may be implemented in a computer program product comprising instructions which, when the program is executed by a computer, cause the computer to perform the acts.

The outlined actions may be implemented in a wind turbine generator control system comprising means for performing the actions. As outlined, the wind turbine generator control system may be implemented using a computer program product and a computer.

In one aspect, a wind turbine generator control system may include an apparatus for measuring a set of blade load values, which may include strain gauge sensors mounted on an inboard side of a blade root surface.

In one aspect, the strain gauge sensors are arranged in a wheatstone bridge configuration.

In the example, the tip clearance reconstruction and IPC hub load inputs are dependent on the root load.

In one aspect, the blade root load may be acquired using strain gauge sensors mounted on the inside of the blade root surface. The controller can obtain three root sensor signals as load inputs and the sensors can be calibrated.

Bracket glue as well as epoxy glue and GRP (glass fibre reinforced plastic) may be used on the inner surface of the blade in order to avoid damaging the blade surface. The sensor can be mounted on the bracket by screws, so that the technical scheme is convenient to maintain.

The strain sensor may consist of a strain gauge, a signal amplifier, and mechanical protection. The strain sensor is arranged to measure the micro-strain of the surface of the blade root by means of a strain gauge wheatstone bridge. The sensor is arranged to convert the micro-strain into a voltage signal representing the calibrated blade root load.

In addition to the actual (closed) control loop system or method, actions and components may be isolated to provide a method or system of estimating tip clearance. Such a method of estimating the tip clearance between the blade tip and the tower of a wind turbine generator may comprise the outlined actions without generating or performing the control actions.

Drawings

The invention is described by way of example only and with reference to the accompanying drawings, in which

FIG. 1 illustrates a Tip Clearance (TC) of a wind turbine generator;

FIG. 2 illustrates a wind turbine generator tip clearance control flow or controller;

FIG. 3 illustrates an example of rotor and blade and tip clearance sampling;

FIG. 4 illustrates tip clearance values over time;

FIG. 5 shows a control system having a first controller and a second controller;

FIG. 6 illustrates the effect of Tip Clearance Control (TCC);

FIG. 7 illustrates a tip clearance control system or method of a wind turbine generator; and

FIG. 8 shows a fit of the data used in estimating tip clearance.

Detailed Description

Fig. 1 shows a wind turbine generator comprising a nacelle 3, a rotor 4 with at least one blade rotating relative to a tower, and a tip gap between a tip 7 of the at least one blade and the tower. The blade 5 extends from a blade root 6 towards a blade tip 7. A Tip Clearance (TC)8 exists between the tip 8 and the tower 2. The Tip Clearance (TC) is the lowermost blade 5 of the rotor 4.

FIG. 2 illustrates a method or system of tip clearance control 1000 of the wind turbine generator 1 as illustrated in FIG. 1. The method comprises the act of measuring 1100 a set of operational values 110 on the wind turbine generator 1. The operational values 110 may include a pitch value, a generator speed value, and a torque value. Operational values 110 include a set of blade load values 150 that may be obtained by, for example, a blade load measuring device and strain gauge sensors 20.

There is the act of estimating a tip clearance 210 between the tip 7 and the tower 2 based on the set of operating values 110 and the set of blade load values 150.

There is an act of generating 1300 control instructions 310 from the estimated tip clearance 210. The act of generating control instructions 310 is performed by a controller 300, here shown as a first controller 320, providing first control instructions 311 to a second controller 340, the second controller 340 generating second control instructions 312.

There is the further action of adjusting 1400 the wind turbine 1 by individually pitching each illustrated blade in accordance with the assumed three individual pitch control commands of a three-blade wind turbine as illustrated in fig. 1.

The first controller 320 and the second controller 340 are described in more detail in fig. 4.

Instructions that may be executed on a computer may be stored in computer program 400.

Fig. 3 shows the rotor 4 and the blades 5 passing through the tower 2 (the centre of which is shown as a dashed line). As the blade passes through the tower 2, the sample estimates the tip clearance over the azimuth range in the sampling window 170, which may cover, for example, 175 and 185 degrees.

For every 120 degrees of azimuth, there will be 110 degrees of azimuth dwell and hold.

For example, FIG. 4 illustrates the actual tip clearance 8 and the estimated tip clearance 210 of the blade 5 to the tower 2. For a particular wind turbine, the X-axis is time in seconds and the Y-axis is meters.

The platform seen in the estimated tip clearance 210 is a sample and hold of each blade 5 spaced 120 degrees apart in the three blade 5 rotor 4 that is maintained at 110 degrees azimuth.

Fig. 5 illustrates a dual PI control structure. The tip clearance set point (TC SP)215 may be a constant parameter or dynamically adjusted. The error between the tip clearance set point 215 and the estimated tip clearance 210 value is converted to a hub pitching moment set point for the first PI controller (PI 1). As shown, the tip clearance set point 215 is compared to an estimated tip clearance 210 provided by the tip clearance estimator 200 (not shown herein). The error or result of this comparison is provided to a first controller 320, which may be a PI controller shown as PI1, the first controller 320 being arranged to output a wind turbine hub pitching moment set point.

The pitch moment setpoint is compared to the pitch moment and the error or result of the comparison is provided to a second controller 340, which may be a PI controller, shown as PI 2, the second controller 340 being arranged to output a pitch control command.

In summary, FIG. 5 illustrates a particular aspect of FIG. 2, namely the act of generating 1300 the control instructions 310 by the act of generating 311 a first instruction based on the estimated tip clearance 210 and the set point tip clearance 215. The first command 311 is a setpoint tilting moment command. This action is followed by an action of generating a second command 312 according to the first command 311 and the measured hub pitching moment, wherein the second command 312 is a pitch command. The second instructions 312 may be further transformed into control instructions 310 (not shown).

FIG. 6 illustrates the effect of Tip Clearance Control (TCC). The upper graph shows pitch angle control with Tip Clearance Control (TCC)1000 (deeper lines) and without tip clearance control (NO TCC) (shallower lines). The lower graph shows the tip clearance 8.

The tip clearance 8 is shown with and without the tip clearance control 1000 method (shallower lines).

If the tip clearance 8 is below the TCC threshold (see around the X-axis 151) for a particular cycle, the TCC will begin operating in some subsequent cycles.

As shown, after the first cycle (around 151) the controller of the wind turbine starts pitching and the tip clearance increases.

FIG. 7 summarizes an embodiment of a Tip Clearance Control (TCC)1000 of the wind turbine 1 to increase the tip clearance. This embodiment may be understood as a series of acts, collecting the system components or acts installed on a computer or in a computer program product, such that when an act is performed in accordance with the instructions, the computer will generate the instructions. With the wind turbine product outlined, TCC (tip clearance control) improves the safety of the tip clearance and also ensures minimum tip clearance and meets critical load requirements.

The closed loop TCC1000 and wind turbine generator 1 system operate as follows.

After obtaining the data from the sensors, the TCC function will enable the IPC (individual pitch control) algorithm to bend the rotor plane of the wind turbine generator 1 to avoid extreme tip clearances.

IPC-SP (set point) is obtained from the set point table.

If the IPC-PV (processed value) is lower than SP, the IPC-PI (proportional integral) controller adjusts PV to SP. The TCC loop will also obtain IPC-SP and compare it to the original IPC-SP. When TCC-SP is greater than IPC-SP, IPC will work even if the hubload is within the normal range.

The TC model 200 is used to simulate tip clearance. Since the tip clearance cannot be used directly as an input to the controller, the turbine performance or operating signals are used to reconstruct the tip clearance. Such operational values may include blade root load, pitch angle, generator speed, torque, and hub tilting moment.

The tip clearance in the TC model 200 can be estimated by solving the following equation.

Y_YC＝x₁+x₂·θ+x₃·ω+x₄·T+x₅·M_{Blade root}+x₆·M_{SH Tilt}+x₇·F_{Push away}

AX＝Y

Y_TCIs the tip clearance. I.e. the distance between the tip of the blade and the tower as the blade passes the tower. Tip clearance may be a similar measure of the distance between the tip of the blade and the tower or stationary component.

The operating values are as follows: θ is the pitch angle, ω is the generator speed, and T is the generator torque.

M_{Blade root}Is a measure of the flapping moment of the blade root. M_{SH Tilt}Is a measure of the tilting moment of the stationary hub. M_{SH Tilt}A non-linear combination of flap loads for 3 different blade roots can be obtained.

F_{Push away}Is the thrust of the rotor. F_{Push away}A linear combination of flap loads for 3 different blade roots can be obtained.

x1 through x7 are constants determined based on the operating data, for example, by regression fitting and a system of equations.

In the example, the constant is determined using a standard regression fit model. Least squares fitting demonstrates a method of approximating a fitted curve.

The X-parameter is determined, and then the estimated tip clearance may be determined as:

Y′＝AX

the estimated value of the tip clearance is thus a least squares fit to the actual value of the tip clearance Y.

The estimated value Y' can be regarded as the actual value Y with the relevant tolerance. In addition, well-functioning estimates are easily implemented as linear regression methods in hardware.

M_{SH Tilt}May be three different M_{Blade root}The result of the nonlinear result of the value is determined as

F_{Push away}Obtaining linear combinations of flap loads for three different blade roots

F_{Push away}＝(M_br1+M_br2+M_br3)/3L

In azimuth, L is the blade length.

Thus, the root load sensor values are used to estimate tip clearance and perform tip clearance control.

Referring to fig. 8, time series data is extracted from the wind turbine as the blade passes the tower in the design load case 1.x (DLC 1. x). The mean data was sampled in azimuth 180 ± 5 degrees. The total data quantity is n, and an original data set is obtained. Data fitting was performed for DLC1.2, DLC1.3, DLC1.4 and DLC1.5 to alter the percentage contribution of the different DLC to the results.

In the above equation, the data is provided in matrix a, and a linear regression can be performed by a least squares fitting method. The parametric contribution is the result of the normalization of a and constants in the vector X. In particular, the highest correlation of root moment to tip clearance has been observed.

The confirmed parameters were used to check the fitting results for DLC 1.x in fig. 8. The emphasis is on the estimated tip clearance of less than 10m, and the reconstruction error is about + -0.5 m.

The data example and azimuth sampling shown is shown in figure 3. The estimated tip clearance signal for each blade is processed into an input tip clearance signal that is provided to the controller. As shown in FIG. 4, the tip clearance signal is highly discrete due to the sampling segment.

As the blade passes the tower, the controller samples the azimuth angle between, for example, 175 and 185 degrees for the estimated tip clearance. For every 120 degrees of azimuth, there will be a 110 azimuth "dwell and hold" state.

Optionally there is a controller, for example a PI controller with two options. There is a first option when the control loop is open-loop control that maintains the estimated tip clearance at the set point. There is a second option to maintain the estimated tip clearance as an output value of the function and use it in closed loop control.

The output of the TCC model is to raise or raise the lower IPC set point, which will produce an IPC effect that bends the rotor plane.

There is a dynamic look-up table. The original IPC was set to process the hub pitching moment and hub yawing moment signals by using the coleman conversion to reduce the rotor asymmetric load. The PI controller may be arranged to adjust the load reduction effect. The reduction of the fatigue load of the hub will cause or accumulate damage to the pitch bearing.

Dynamic lookup tables allow the IPC function to benefit not only from extreme loads.

The enable and disable conditions are look-up tables, which are also control set points. The lookup table has an upper limit and a lower limit.

If the hub torque handling value is outside the lookup table range, the IPC function will function to control the hub torque within the table limits. This will (significantly) reduce unnecessary IPC pitching behaviour and thereby reduce pitch bearing damage.

The lookup table is a two-dimensional table including wind speed and corrective hub torque values. The wind speed data comes from a wind estimator in the controller and there may be specified values that depend on the trade-off between hub fatigue loading and pitch bearing damage.

The TCC function only changes the IPC set point lower limit to control hub bending to avoid extreme tip clearances. Therefore, normal IPC and TCC do not conflict.

Thus, sharing a single IPC set point table (IPC-SP) and attempting to keep the hubload within normal ranges is achieved.

For load prediction, to improve IPC capability under extreme wind shear conditions and extreme wind change conditions, the rate of change of hub load is also considered. The rate of change of the filtered hubload will change the dynamic IPC lookup table. If the module detects a rapid change in hubload, then the IPC setpoint lower limit will be raised or the upper limit lowered to predict IPC behavior.