KR101706508B1 - Data Analysis System of Mechanical Load Measurement Data for Wind Turbine - Google Patents

Abstract

The present invention relates to a system for analyzing fatigue and an equivalent load of a wind power generator capable of determining the suitability of a fatigue analysis performed by a manufacturer of a wind power generating system during a designing process by installing a strain gauge, a vibration sensor, and the like on parts such as impellers, a shaft, a nacelle, a tower, and the like which are major parts of the wind power generator, measuring a bending moment, a torsional moment, vibration, and the like generated during the operation of the wind power generating system, and performing the fatigue analysis based on the measurement. To this end, a system for analyzing fatigue and an equivalent load of a wind power generator according to an embodiment of the present invention includes: a specimen of the wind power generator; a sensor unit measuring a physical quantity of the specimen; and a control unit calculating time-series data and statistical data based on the physical quantity measured by the sensor unit, and analyzing the fatigue and the equivalent load through a rainflow counting method.

Description

Technical Field [0001] The present invention relates to a fatigue analysis method and an equivalent load analysis system for a wind turbine generator,

The present invention relates to a fatigue analysis and equivalent load analysis system for a wind turbine. More particularly, the present invention relates to a method of fatigue analysis and equivalent of a wind power generator that performs time series and statistical value analysis, fatigue analysis and equivalent load analysis through various stages of data processing of vast amounts of physical quantities measured in main parts such as wind turbine blades, And a load analysis system.

Wind power is a technology that uses a mechanical energy to generate electricity by rotating rotor by using aerodynamic characteristic of kinetic energy of air flow.

A wind turbine using such wind power is composed of a rotor composed of a blade and a hub as main components, and a generator driven by rotation and located inside Nacelle.

Since these wind generators use infinite winds without causing environmental pollution, interest in the environment is increasing worldwide, and attention is increasing as a generator of electric energy these days.

Wind power generation is divided into constant speed operation and variable speed operation depending on the operation mode. Among them, the constant speed operation is a method of rotating the turbine at a constant speed irrespective of the change of the wind speed, but there is no need to control the turbine speed, but energy conversion efficiency is low at other wind speeds deviating from the design wind speed. Especially, in regions where wind velocity is not constant like Korea, it is always low efficiency, so it is not used gradually in large capacity wind power generation system.

On the other hand, because the maximum power point changes according to the wind speed, the variable speed operation type wind turbine is pitch controlled to operate the turbine speed to the optimum blade ratio (ratio of the blade end point speed to the wind speed) It is the way to get it.

However, in the case of the conventional pitch control method, the range is lower than the rated operating wind speed and is higher than the rated wind speed, so that the wind power generator is likely to be exposed to an environment where the wind power generator is to be operated.

Especially, it is likely that excessive force is transmitted to the generator side of the wind power generator because it is likely to occur in mountainous areas, typhoons, and gusts like Korea.

In addition, when converted to the rated wind speed range due to sudden gusts, the aero torque of the wind turbine (i.e., the torque of the rotor) is suddenly increased and excessive force is transmitted to the generator side of the wind turbine. At this time, due to the excessive rotation of the rotor, a rotor load is generated, which damages the drive train and the shaft inside the wind turbine.

Japanese Patent Publication No. 2011-501172

In the present invention, a strain gauge, a vibration sensor, or the like is installed on a wing, a shaft, a nacelle and a tower, which are the main parts of a wind turbine, to measure bending and torsional moments and vibrations generated during operation of the wind turbine, And a fatigue analysis and equivalent load analysis system of a wind turbine capable of judging the suitability of the fatigue analysis performed by the wind turbine generator company during the design process.

It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are not intended to limit the invention to the precise form disclosed. It can be understood.

The system for analyzing fatigue and equivalent load of a wind turbine according to an embodiment of the present invention derives time series and statistical data through a sample of a wind turbine generator, a sensor unit for measuring a physical quantity of the sample, and a measured physical quantity of the sensor unit, And a controller for performing fatigue analysis and equivalent load analysis through flow counting.

The sensor unit may include at least one of a strain gage, a displacement gauge, a load cell, a torque, an accelerometer, and a speedometer.

In addition, the installation position of the sensor unit may satisfy the following conditions.

- location where high strain on unit load occurs, position that can avoid load inflow path, position where uniform stress is measured (that is, place where high stress / strain gradient is not taken , Position where stress concentration is not applied), position to ensure sufficient space, location where temperature can easily be determined, position on material with uniform load, position of material that can easily fix or attach measuring equipment.

The validity check of the measured physical quantity in the sensor unit may be implemented in at least one of the following data conditions.

- Measurement error data caused by calibration of the sensor part, operation range of the sensor part, noise, etc., exclusion data of the measurement part which is out of the measuring range of the sensor part and the data collecting device, reference data measured in the normal state, Verification of suitability through real-time mutual comparison of data measured in the sample Data measured in conditions such as wind speed distortion or wake effect caused by data, the terrain near the sample or other wind power generators or buildings, Data measured during the course of the yawing of the nacelle due to sudden changes in the wind direction.

Also, the equivalent load is represented by a value at a frequency of 1 Hz and can be expressed by the following equation.

[Mathematical Expression]

(Where R _eq is the equivalent load, Ri is the i th load, ni is the number of the i th load, n _eq is the number of cycles equivalent, and m is the SN curve slope of the material)

The method for fatigue analysis and equivalent load analysis of a wind turbine according to the present invention includes a step I for measuring the load of a sample of a wind turbine, a step II for measuring a physical quantity of the sample, and a step for checking the validity of the measured physical quantity, And the step (III) of deriving the fatigue analysis and the equivalent load analysis of the sample.

The present invention provides a fatigue analysis and equivalent load analysis system of a wind power generator that automatically performs time series and statistical values, fatigue analysis, and equivalent load analysis using the physical quantity of a wind turbine generator, The data is read and fatigue analysis is performed.

In addition, the fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention can be realized by installing strain gauges and vibration sensors on the main parts of a wind turbine such as wings, shafts, nacelles and towers, It is possible to judge whether the manufacturer of the wind power generation system is suitable for the fatigue analysis which has been carried out during the design process by performing the fatigue analysis based on the measured bending and torsional moments and vibrations.

It should be understood, however, that the effects obtained by the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those skilled in the art to which the present invention belongs It will be possible.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate preferred embodiments of the invention and, together with the description, And shall not be interpreted.
Fig. 1 shows a configuration diagram of a fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention.
2 is a block diagram of a data acquisition device (IMC module) in a fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention.
FIGS. 3A to 3C are graphs showing time-series data of wind speed, output, rotor rotational speed, and mechanical load measurement values during a performance test period in the fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention, Is a graph showing results of analysis of measured physical quantities generated at main portions of a wind turbine relative to wind speed in a fatigue analysis and equivalent load analysis system of a wind turbine according to an embodiment of the present invention.
FIG. 4 is a graph showing the natural frequency measurement results according to the states of the wind turbine generator, such as start, stop, yaw, etc., in the fatigue analysis and equivalent load analysis system of the wind turbine according to the present invention.
5A to 5B are graphs showing the cumulative number of fatigue load spectra of each 10-minute data obtained by the rain flow counting in the fatigue analysis and equivalent load analysis system of the wind turbine according to the present invention, The equivalent load expressed by 1 Hz is obtained by using the load obtained by the fatigue analysis and equivalent load analysis system of the wind turbine and the cumulative frequency.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The embodiments of the present invention are described in order to more fully explain the present invention to those skilled in the art, and the following embodiments may be modified into various other forms, The present invention is not limited to the embodiment. Rather, these embodiments are provided so that this disclosure will be more faithful and complete, and will fully convey the scope of the invention to those skilled in the art.

Also, in the following drawings, each configuration is exaggerated for convenience and clarity of description, and the same reference numerals denote the same elements in the drawings. As used herein, the term "and / or" includes any and all combinations of any of the listed items.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

As used herein, the singular forms "a," "an," and "the" include singular forms unless the context clearly dictates otherwise. Also, " comprise "and / or" comprising "when used herein should be interpreted as specifying the presence of stated shapes, numbers, steps, operations, elements, elements, and / And does not preclude the presence or addition of one or more other features, integers, operations, elements, elements, and / or groups.

The present invention specifies "fatigue analysis and equivalent load analysis" which is one of the most important performance test items of wind power generation system. It includes strain gauges and vibration sensors at the main parts of wind turbine system such as wing, shaft, nacelle and tower The wind power generation system is designed to measure the bending and torsional moments and vibrations generated during the operation of the wind power generation system, and then the fatigue analysis is performed based on the measurements to determine whether the fatigue analysis is appropriate to be.

This physical measurement requires a continuous measurement from the starting wind speed to the end wind speed section based on the state of the wind turbine generator in the natural standby state (power generation, idling, emergency stop, etc.) Therefore, the present invention intends to describe a detailed test procedure for the measurement examination as an alternative, in the case of the mechanical load test of the wind power system, as repetitive measurements under the same conditions, i.e., comparative tests, are virtually impossible.

A strain gauge and a vibration sensor are attached to the wing, main shaft and tower parts of the wind power generation system to convert the amount of displacement measured during operation of the wind power generation system into a load value. Then, fatigue analysis and equivalent load analysis are performed. Applicable sensors and data collection devices are pre-calibrated and measured by a national / international calibration laboratory before data collection is performed.

The fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention comprises a sample (wind power generation system), a strain gauge, a vibration meter, and a measurement module (IMC).

Here, in the embodiment of the present invention, the sample used for the measurement is a 1.5 MW wind power generation system (Model: HJWT-1500), and the measurement result is described.

First, the fatigue analysis and equivalent load analysis system of the wind turbine perform the examination of the test site and the appropriateness examination of the test equipment.

In the examination of the test site, it was confirmed that the meteorological tower installed on the test site of the mechanical load measurement was installed to satisfy the following requirements on the wind turbine to measure the wind speed acting on the wind turbine. The influence of other wind power generation systems in the vicinity is excluded, and an effective measurement bearing section is calculated.

In addition, check that equipment such as strain gauges, vibration systems, mechanical physical quantity collection devices required by the test standards meet the specifications required by the following standards.

Then, at the start of the test, the physical quantities to be measured are examined as follows, the requirements of the sensor section are reviewed, data is collected, data is sorted, the measured data are verified, The final result is obtained.

① Measured object physical quantity

○ Basic weight of wind power generator

- Blade: flap and edge bending moment

- rotor: tilt moment, yaw moment, torque,

- Tower: Bending moment

○ meteorological quantity

- Wind speed: Based on hub height

- Wind direction: Based on hub height

○ Operation quantity of wind power generator: It is possible to utilize wind power generator controller data.

- Power: Mandatory

- Rotor speed: duty

- Pitch angle: Applicable only to obligatory variable pitch wind turbine generator

- Yoying Location: Mandatory

- Rotor azimuth: duty

- Grid connection: recommended

- Brake condition: recommended

② Sensor type

○ Strain gauge bridge

○ Load cell / torque tube (including piezoelectric cell)

○ Accelerometer, speed, rotation and displacement transducer

③ Determination of sensor position

○ The position of the sensor that meets the following conditions is selected as the attachment position of the sensor for measuring the total structural load.

- location where high strain on unit load occurs

- a point that provides a linear relationship between stress and load and avoids the load inflow path

- where there is a uniform stress (ie, where there is no high stress / strain gradient, where stress concentration is not)

- enough space to attach the sensor

- Where temperature can be easily determined

- An object with a uniform load (for example, a single iron is better than a composite material)

- materials that can be easily fixed or attached to measuring equipment

(2) Physical quantity collection: wind turbine condition standard

Measurement load cases (MLC) of steady-state operation

- Power production

- Power production with occurrence of fault.

- No output: parking, idling,

Measurement load cases (MLC) of transient events

- start up

- normal shut-down

- Emergency shut-down

- grid failure

(3) Physical quantity collection matrix: Time series data based on mean wind speed and turbulence intensity

○ Power generation conditions

- The number of data accumulated in the 10-minute time series in each wind speed section up to the rated wind speed (v _r ) should be at least 30 or more, by dividing the wind speed to 1 m / s and the turbulence intensity to 2% , Measurements should be recorded at various different turbulence intensities. At least four turbulent sections in each wind speed section should contain at least three time series.

- The cumulative number of time series for 10 minutes in each wind speed section should be at least 8 in the range of wind velocity lower than 5 m / s lower than the terminal wind speed v _out at rated wind speed v _r . However, there are no additional conditions for the turbulence intensity in the section.

Termination velocity v of the time series of time to be recorded in the interval of from 5 m / s lower than the wind speed to the _end-out wind speed v _out may be reduced to two minutes.

Termination velocity v in each velocity interval of from 5 m / s lower than the wind speed to the _end-out wind speed v _out than 1 m / s low wind speeds to be recorded at least three time-series.

At least one time series shall be recorded in the termination wind speed v _out .

Termination velocity v from 5 m / s wind speed lower than the _end-out velocity v in the region of up _out there is no criteria for the turbulence intensity. The 2-minute time series can be obtained from the 10-minute time series. However, there should be no overlap between the 2-minute time series divided by the 10-minute time series.

○ Generation of power generation and error condition

- wind speed rated wind speed v _r than 6 m / s from the low wind speed 2 m / s or higher than the rated wind speed low wind speed, the rated wind speed than 2 m / s from the low wind speed 2 m / s than the rated wind speed high wind speed and the rated wind speed 2 It is divided into three sections above m / s high wind speed. The duration of each time series should be at least 2 minutes.

○ No output status (stop or idle)

- The wind speed range for stop or idling measurement load cases (MLC) is recommended to be 4 m / s. The duration of the time series is recommended to be 10 minutes. Measurements in the no-output state should be made at misalignment angles of various yaws, including the least undesirable flow inlet angle.

(4) Verification of measurement data

Data validation: data validation

- Measurement error caused by data calibration, sensor operating range, noise, etc.

- Exclusion from measurement results beyond the measurement range of sensors and data collection equipment.

- Verification of compliance through real-time mutual comparison between reference data measured in normal condition and data measured in wind turbine under test.

- wind direction data measured under conditions such as the wake effect caused by other objects installed near the test wind turbine

- Data measured during yawing due to sudden changes in wind direction during data measurement.

(5) Analysis of measurement data

Time series data (time series)

- All measured loads and statistics shall be indicated for the average wind speed and the average, maximum, minimum and standard deviation shall be indicated.

○ Load spectrum

- The data calculates the turbulence intensity through high-pass filtering after only 10 minutes of sampling at normal average wind speed.

○ Rainflow counting

- The fatigue load spectrum for each measured time history is calculated using the re-flow counting method.

- Each load range should be divided into at least 50 and re-flow counting.

○ Equivalent loads

- Equivalent load is generally expressed as a value at a frequency of 1 Hz and can be expressed by the following Equation (1).

[Equation 1]

(Where R _eq is the equivalent load, Ri is the i th load, ni is the number of the i th load, n _eq is the number of cycles equivalent, and m is the SN curve slope of the material)

(6) Test results

The results of the test are derived from the following conditions.

① Calibration

- Sensor calibration, measurement system calibration, load calibration

- Summary of Signal Conversion

② Database

- collection matrix: set value for collection matrix - wind velocity, turbulence interval, wind direction range, etc.

- Data validation: data sorting and tolerance limits, description of removed data

- Final database: the specification, the number of files, etc. for the remaining data after the feasibility check.

③ Time response and load statistics

- City of time response for all relevant channels at different wind speeds

- Urban representation of the summary of load statistics (see Annex C)

④ Wind power generator dynamics

- Interpretation method: description of methodology and algorithm - resolution, average, weight

- the frequency spectrum of the off-line wind turbine: a note for all relevant channels

City of waveness spectrum

- The frequency spectrum of the wind turbine in operation: the frequency spectrum of all relevant channels at low and high wind speeds.

⑤ Fatigue load

- Interpretation method: description of methodology and algorithm - resolution, average, weight

- Load spectrum: City of the load spectrum for all relevant channels. Record various tables

- Equivalent load: For a given turbulence and yawing error,

Graph of Equivalent Load vs. Wind Velocity

⑥ Uncertainty assessment

- calibration uncertainty

- Uncertainty of measurement load, time series, average load, equivalent range and spectrum shall be indicated to three decimal places. Effective number of measurement uncertainty shall be calculated as two or three digits. .

The results of the mechanical load measurement test through the fatigue analysis and equivalent load analysis system of the wind turbine of the present invention will be described below. The mechanical load measurement test results to be described later show the results according to one embodiment, and thus the present invention is not limited to these embodiments.

1. Introduction

Analyzes the mechanical load applied to the wind turbine to determine the mechanical stability of the wind turbine.

2. Wind generator

○ Name: 1 500 kW Wind Power System Usage: Grid type

○ Model: Horizontal axis type steel tower structure: Circular tube type

3. Measuring place

○ 1,500 kW wind turbine system installed at the Korea Renewable Energy Research Center (294)

4. Using the device

4.1 Configuration diagram of measuring device

1 500 kW The arrangement of the mechanical load measuring devices installed in the wind power generation system is as shown in Fig.

4.2 Data acquisition device (IMC module)

The data acquisition device is configured as shown in FIG.

○ BusDAQ

Ethernet (100 Mbit)

Storage on PC / network drive

CAN max. Baud rate 1 MBit / s

○ Uni8

Max Sampling rate 1 000 Hz

Max Bandwidth 200 Hz

8 Channels

Full Bridge / Half Bridge / Quarter Bridge / DC-Excitation

○ Bridge2

Max Sampling rate 5,000 Hz

Max Bandwidth 1,000 Hz

2 Channels

Full Bridge / Half Bridge / Quarter Bridge / DC-Excitation

○ C8

Max Sampling rate 100 Hz

Max Bandwidth 20 Hz

8 Channels

Voltage up to 50/60 V

5. The measured values required for mechanical load are as follows.

- Vibration of the nacelle: Accelerometer measures the horizontal and vertical acceleration by attaching an accelerometer to the frame inside the nacelle.

- Rotor angular speed: Calculated by using generator speed and gear ratio.

- Rotor azimuth angle: Calculated using the edge moment of the wing.

- Blade loading: Measurement of flap direction and edge direction moment on three wings.

- Tower Moment: Fixed on the surface of the earth, requiring consideration of the yaw angle for alignment with the nacelle coordinate system.

6. Time response and load statistics

FIGS. 3A to 3C are graphs showing time-series data of wind speed, output, rotor rotational speed, and mechanical load measurement values during a performance test in the fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention. Minute, wind speed, output, and rotation speed are represented by 10 minutes data.

FIGS. 3D to 3F are graphs showing the analysis results of the measured physical quantities generated at the main parts of the wind turbine relative to the wind speed in the fatigue analysis and equivalent load analysis system of the wind turbine according to the embodiment of the present invention.

First, the load applied to the blade increases greatly in the flap direction as the wind speed increases, while the amplitude in the edge direction load increases as the wind speed increases, instead of being located near the average value of "0".

On the other hand, in the case of turbulence, the output is increased while the wind speed is decreased, but the output is increased, and the output limit is not observed because the measurement of the high wind speed section is not performed. However, The rotor speed increases.

In the case of the bending moment, the load at the bottom of the tower does not change significantly in the vertical direction, but the shape is proportional to the wind speed, and the torque of the main shaft also increases with the wind speed.

7. Wind power generator dynamics

FIG. 4 is a graph showing the natural frequency measurement results according to the states of the wind turbine generator, such as start, stop, yaw, etc., in the fatigue analysis and equivalent load analysis system of the wind turbine according to the present invention.

The natural frequency measurement according to the condition of the wind turbine generator, ie, start, stop, yawing, etc. is performed by measuring the vibration generated in the nacelle at the top of the wind turbine generator by installing an acceleration sensor in the nacelle, For FFT analysis, a sufficient length of data is required, and in order to observe the frequency spectrum widely, it is determined according to the number of sampled data. In the present invention, the data was measured at 50 Hz and the analysis was carried out. The results are shown in FIG.

8. Fatigue load

For fatigue measurement, the equivalent load is generally expressed as a value at a frequency of 1 Hz and expressed by the following equation (1).

[Equation 1]

(Where R _eq is the equivalent load, Ri is the i th load, ni is the number of the i th load, n _eq is the number of cycles equivalent, and m is the SN curve slope of the material)

In addition, the accumulated number of 10-minute data for each load subjected to the re-flow counting is shown in FIGS. 5A and 5B. 5A to 5B are graphs showing a cumulative number of fatigue load spectra of each 10-minute data obtained by the rain flow counting in the fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention.

Also, the equivalent load represented by 1 Hz is obtained by using Equivalent Load and Cumulative Number of Equations (1) described above, and the equivalent load is different depending on the load index (m) 5c to 5e, the equivalent load tends to increase as the wind speed increases for all the parts.

5C to 5E are graphs showing results obtained by calculating the equivalent load expressed by 1 Hz using the load and the cumulative number of times obtained by the fatigue analysis and equivalent load analysis system of the wind turbine according to the present invention.

In addition, the fatigue analysis and equivalent load analysis system of the wind turbine according to the present invention can increase the confidence level by estimating the measurement uncertainty based on the measurement uncertainty calculation basis and reflecting the measurement uncertainty on the measurement result.

In the fatigue analysis and equivalent load analysis system of a wind turbine according to the present invention, the calculation uncertainty of measurement is as follows.

1. Wind speed measurement uncertainty

1.1. summary

1) Estimate the measurement uncertainty of the wind speed for the wind power system.

2) Usage standard KS C IEC 61400-13 Wind power generation system - Part 13 Mechanical load measurement

3) For large-scale horizontal axis wind turbine with 40m ² or more flow area, stipulated for mechanical load measurement of wind turbine generator.

1.2. Uncertainty modeling

)은 측정시스템에서 Display 되는 지시 값을 읽어서 나타낸다.Wind speed

) Indicates the indication displayed on the measuring system by reading it.

: 풍력발전기에 인가되는 풍속 (m/s)

: Wind speed applied to wind turbine generator (m / s)

: 풍력발전기 측정시스템에 표시되는 지시 값 (m/s)

: Indication value (m / s) displayed on the wind turbine measuring system

1.3. Estimation of measurement uncertainty

1.3.1. Wind speed measurement data

- uncertainty estimation type A type

The standard deviation of 5 repeated measurements is 0.14 m / s

)Standard uncertainty by repeated measurements (

)

0.19 / 5 = 0.09 m / s

1.3.2. Measuring instrument data

- uncertainty estimation type B type

(1) Calibration certificate

The calibration certificate extended uncertainty is 0.1 m / s (confidence level approx. 95%, coverage factor k = 2)

Degree of freedom 8

),Therefore, the standard uncertainty due to equipment calibration (

),

0.11 / 2 = 0.06 m / s

(2) Anemometer operating characteristics

이다The standard uncertainty of the anemometer operation is as follows. Considering this as a rectangular distribution,

to be

Degree of freedom 8

)Therefore, the standard uncertainty of flow distortion due to the installation of an anemometer

)

(k=풍속계 등급번호, 1.2, 여기서 k =1)

(k = anemometer grade number, 1.2, where k = 1)

 (3) Flow distortion due to anemometer mounting

이다The standard uncertainty of flow distortion due to anemometer mounting is estimated to be ± 1% of wind speed. Considering this as a rectangular distribution,

to be

Degree of freedom 8

)Therefore, the flow-distortion standard uncertainty due to the installation of an anemometer (

)

(4) Data processing device

이다It is within ± 0.1% of DLS (Meteorological Data Logging Sub System) that receives and processes the signal of the speed sensor. Therefore, considering this as a rectangular distribution,

to be

Degree of freedom 8

)Therefore, the standard uncertainty (DLS)

)

1.3.3. Sensitivity coefficient

According to the modeling, the measured value is a linear function, and the sensitivity coefficient when partial differentiation is 1. Also, since all uncertainty factors affect independently, the sensitivity coefficient, 1, which is calculated as the partial derivative, applies.

1.3.4. Composite standard uncertainty

All the uncertainty factors are multiplied by the positive square root of the uncertainty propagation law to obtain the combined standard uncertainty (the sensitivity coefficient is 1, so it is not indicated separately)

1.3.5. Effective degree of freedom

The effective degree of freedom is obtained by applying the Welch-Satterthwaite formula as follows.

1.3.6. Inclusion factor

In the t distribution table, the confidence level is about 95%, the effective factor of freedom is 8, and the coverage factor k is 2.

)1.3.7. Extended uncertainty (

)

The expanded uncertainty is determined by multiplying the composite standard uncertainty by the inclusion factor.

Therefore,

2 X 0.16 = 0.32 m / s

In case of 2 digits, the measurement uncertainty is as follows.

0.32 m / s (confidence level about 95%, k = 2)

1.4. Report the measurement

The representation of the measurement values reflecting the measurement uncertainty is as follows

Measured value ± Extended uncertainty (confidence level approx. 95%, k = 2)

(12.42 ± 0.32) m / s (confidence level about 95%, k = 2)

1.5. Uncertainty summary table

2. Uncertainty of blade direction and blade direction measurement

2.1. summary

1) Estimate the measurement uncertainty of the equivalent load for the wind power system.

2) Specifications: KS C IEC 61400-13 Wind power generation system - Part 13 Mechanical load measurement

3) For large-scale horizontal axis wind turbine with 40m ² or more swim area, which is specified for mechanical load measurement of wind turbine generator.

2.2. Uncertainty modeling

The equivalent load is defined by the following equation.

From here,

R _eq : Equivalent load (Nm)

R _i : load corresponding to the i-th class of the fatigue load spectrum

n _i : the number of cycles corresponding to the i-th class of the fatigue load spectrum

n _eq : equivalent cycle number

m: slope of the S-N curve for the relevant material

2.3. Estimation of measurement uncertainty

The uncertainty of measurement in the direction of the smoothing (f) and the direction of the blade (e) consists of the following factors.

<Uncertainty factor>

2.3.1 Calibration uncertainty

The uncertainty of the constant A for the relationship between the sensor and the bending moment.

Measure the electrical signal generated by the actual load and calculate the uncertainty of the constant A by regression analysis.

1) Determination of the calibration matrix

The flexural moment of the rotor in the smoothing direction (f) and the blade direction (e) are measured using a strain gauge in the smoothing direction (f) and a strain gauge in the blade direction (e). The dependence of the measurement signal on the bending moment (M) in the smoothing direction (f) and the blade direction (e) is determined by calibration.

The possibility of interference effects of the bending moment of the blade direction e with respect to the signal of the smoothing direction f and the interference effects of the bending moment of the smooth direction f with respect to the signal of the blade direction e are also And shall be amended during the load measurement process.

This section estimates the calibration factor A through actual measurements. The unit of A is the sensitivity coefficient of the strain gage to the bending moment (μV / V / Nm).

The slope values from the measurement matrix are as follows.

From here

：평활 방향(f)의 스트레인게이지 브리지의 신호(μV/V)

: Signal (μV / V) of the strain gauge bridge in the smoothing direction (f)

：날 방향(e)의 스트레인게이지의 브리지의 신호(μV/V)

: The signal (μV / V) of the strain gage bridge in the blade direction (e)

：평활 방향(f)의 굽힘 모멘트(Nm)

: The bending moment Nm in the smoothing direction f,

：날 방향(e)의 굽힘 모멘트(Nm)

: The bending moment Nm in the blade direction e,

Ai (i = 1 ... 4): Measurement factor. The A2 and A3 constants are caused by interference and are ideally zero. The value of the calibration factor A and the associated uncertainty of type A (s _Ai ) are derived by linear regression using measurement data.

Y = ax + b

<Standard uncertainty obtained through regression analysis>

The unit of calibration factor A is (μV / V / Nm), but in practice, mV / V / kNm is used.

The type B uncertainty (u _Ai ) of the calibration factor A has the following uncertainty factors.

와

)- measured signal (

Wow

)

- the measured loads (uF _f and uF _e )

- the measured distance between the strain gage and the working load (uL _f and uL _e )

- the direction of the load acting on the rotor axes (u _af and u _ae )

- the direction of the load acting on the rotor protrusions (u _af and u _ae )

This uncertainty is determined by experimentation as the maximum applied calibration load.

<Calibration load main measurement value>

The uncertainty of type B at each calibration coefficient is determined by the following method.

(Unit of A: mV / V / Nm)

The formula of A1, A2, A3, A4 consists of five factors, and the sensitivity coefficient is obtained by partial differentiation for each factor.

Assuming there is no correlation between the uncertainties of type B:

을 각 인자별로 편미분한다.

Are differentiated by each factor.

The values of the parameters used in this case are shown in the following table.

That is, cosa and cos &amp;thetas;

Apply the sensitivity coefficient and the uncertainty propagation rule by the individual factors obtained by the partial differentials to obtain the uncertainty of A1 with the square root of the positive.

도 적용한다.In the same way

.

를 인자별로 편미분한다.

.

Apply the sensitivity coefficient and the uncertainty propagation rule by the individual factors obtained by the partial differentials to obtain the uncertainty of A2 with the square root of the positive.

를 인자별로 편미분한다.

.

The uncertainty of A3 is obtained by applying the sensitivity coefficient and the uncertainty propagation law by the individual factors obtained by the partial differential to the square root of the positive.

를 인자별로 편미분한다.

.

Apply the sensitivity coefficient and the uncertainty propagation rule by the individual factors obtained by the partial differentials to obtain the uncertainty of A4 with the square root of the positive.

) B형 (

)의 불확도를 불확도 전파법칙을 적용하여 합성하여 계산된다.The uncertainty for the calibration factor A is the uncertainty type A

) Type B

) Is calculated by applying the uncertainty propagation law.

In general, the uncertainty indication should be ua1, but the indication of SA1 is indicated as S _A1 to avoid confusion with the uncertainty of the calibration coefficient A1 according to the expression of IEC TS 61400-13.

2) Use of calibration matrix

a) A non-negligible cross-correlation matrix is used to convert the measurement signal into a load signal using the following inverse calibration matrix.

The uncertainty in the D coefficient can be determined by the values A and u _cA of the above equation.

The sensitivity coefficient is obtained by dividing by the factor of A _i in D ₁ .

D ₂ , D ₃ , and D ₄ are similar forms, and are partial derivatives like the above equations.

The sensitivity coefficient is obtained by dividing the factor of A _i in D ₂ by

The sensitivity coefficient is obtained by dividing by the factor of A _i in D ₃

The sensitivity coefficient is obtained by dividing by the factor of A _i in D ₄ , as follows.

The uncertainty of the coefficient D _i is converted to the measurement uncertainty due to the calibration uncertainty (u _{cal, f} and u _{cal, e} ).

It can be calculated by obtaining both the smoothing direction (f) and the blade direction (e) and A ₂ and A ₄ .

If the interference can be ignored, the use of the calibration matrix is greatly simplified. In this case, the relative uncertainty on the measured load due to the uncertainty of the calibration can be expressed as the relative uncertainty of the calibration coefficient.

2.3.2 Time series

The uncertainty of the value of the sample at each time of the bending moments in the direction of the smoothing (f) and the direction of the blade (e) depends on the uncertainties due to the applied calibration results (u _{cal, f} and u _{cal, e} ) Of uncertainty.

a) Calibration uncertainty

Calibration uncertainty u _{cal, f} And u _{cal, e} are determined by the samples per hour.

b) signal uncertainty

The measured signal uncertainty arises from the noise of the signal (eg electromagnetic contact), zero drift (eg temperature effect of the strain gage), where the D coefficient is calculated from the following formula.

c) Overall uncertainty

The uncertainty of the measurement and the uncertainty of the signal due to the time recording by measuring the bending moments in the direction of the direction of deflection (f) and of the direction of the edges (e) are calculated by the following formula:

2.3.3 Average load on wind speed

Smooth direction (f) Average moment and blade direction (e) The average moment is measured as a function of the mean value measured in 10 minutes of the smooth direction (f) signal, the blade direction (e) signal and wind speed.

a) Calibration uncertainty

The uncertainty of the 10-minute average load due to the calibration uncertainty is determined as follows.

Where N is the whole number in the bin.

b) signal uncertainty

The uncertainty of the 10-minute average load due to the signal uncertainty is determined by the uncertainty of the type A due to the calculation of the standard deviation of the 10-minute average of the considered bin.

c) uncertainty of wind speed

The measured wind speed uncertainty (u _{v, i} ) per bin is determined by the type B uncertainty of the wind speed in Annex D of IEC 61400-12. The slope of the load curve is then used to calculate the relative uncertainty of the bin average bending moment.

Here, i is the index of the bin considered and starts from the second bin (i = 2)

In this case, the change of the load due to the change of the wind speed is calculated for each bin, which is the measurement uncertainty in the bin.

The measurement uncertainty is 0.32 m / s and the standard uncertainty is 0.16 m / s at the wind velocity of 12 m / s.

The load variation at the wind speeds of 11 m / s and 12 m / s is as follows.

uv, e are obtained in the same way as uv, f.

d) Overall uncertainty

The overall uncertainty is calculated as the load range corresponding to each bin average bending moment using the following equation:

At this time, if the uncertainties constituting Uf are 0.269, 0.100, and 0.026 (kNm), the Uf synthesis standard uncertainty is 0.288 kNm.

Also, if the uncertainties constituting Ue are 28.072, 4.888, and 2.379 (kNm), then the Uc synthesis standard uncertainty is 28.592 kNm.

2.3.4 Equivalent load range according to wind speed

The equivalent load to the moments of the smoothing direction (f) and the direction of the blade (e) was measured by measuring the change in signal of the direction of the smoothing direction (f) and the signal of the blade direction (e) It is determined as a function of wind speed. The 10-minute load signal is converted to a time series value of 10 minutes in the direction of the smoothing (f) and the direction of the blade (e) using the measurement matrix (see Equation 5). The time series load for 10 minutes is converted to the load range by the rain flow counting technique.

a) Calibration uncertainty

The uncertainty per equivalent load range of 10 minutes according to the calibration uncertainty is determined by the Monte Carlo method. In this process, the calculation of each equivalent load Req value is repeated a number of times (for example, 10 times) together with the modified time series of the load. The modified time series shall be as follows.

From here,

Mf * (t) and Me * (t): modified time series

Ran: Any number from a normal distribution with an average value of 0 and a standard deviation of 1 [N (0,1)].

ucal, f and ucal, e: 3.2 Calibration uncertainty given in time series

<Assumed equivalent load (kNm) using data used at calibration>

로 표기된다.

Respectively.

The random number, Ran, appears one time in the time series correction, and within ten minutes all time samples are modified using the same value of Ran. In this method, a number of R _eq values (R _{eq *} ) are obtained from the shear direction (f) and the nose direction (e) load for each 10 minute time series.

Where N is a number considered as a bin section.

b) signal uncertainty

The uncertainty of the average equivalent load range for each 10 minutes due to the signal uncertainty is determined by the uncertainty of the A type by calculating the standard deviation of the 10-minute average of the bin intervals considered. It is assumed that the uncertainty of the average 10-minute load value due to the uncertainty of the signal is not correlated. Therefore, the uncertainty of the bin-interval load due to the uncertainty of the signal can be calculated by using the standard deviation of the 10-minute average of each bin interval (stdevReq, f and stdevReq, e) .

c) wind speed uncertainty

(Uv, i) The uncertainty of the measured wind speed is determined by the type B uncertainty of the wind speed in Annex D of IEC 61400-12. This uncertainty is converted to the uncertainty associated with the bin-averaged bending moments (uv, f and uv, e) using the curve of the Req curve:

Here, i is the index of the bin considered, starting from the second bin (i = 2).

The values of uv, f (1) and uv, e (1) are the same as uv, f (2) and uv, e (2), respectively.

Req, f The equivalent wind loads at 90 kNm and 85 kNm are 12 m / s and 11 m / s, respectively. The measurement uncertainty is 0.32 m / s and the standard uncertainty is 0.16 m / s at the wind velocity of 12 m / s. The equivalent load variation at the wind speeds of 11 m / s and 12 m / s is as follows.

uv, e are obtained in the same way as uv, f.

d) Overall uncertainty

The overall uncertainty is calculated for each bin average equivalent load using the following equation:

If the uncertainty of each factor is 3.8, 1.7, and 0.8 kNm, then the total uncertainty Uf is 4.2.

Ue is obtained in the same way as Uf.

If the uncertainty of each factor is 22.3, 9.9 and 3.2 kNm, the total uncertainty Ue is 24.7.

2.3.5 Annual Load Range

The annual load range can be determined by the selected annual wind speed, the selected turbulence intensity, and the number of operational events (grid damage, braking action). This is accomplished through a selected 10-minute time-series rainflow counting by adding the corresponding specific wind conditions and corresponding operating events and the number of occurrences of the number of occurrences per year. The continuous measurement period used for the entire 10 minutes is as follows.

From here,

Ran: random number used as a sample for all 10 minutes of all time

(An arbitrary number extracted from the normal distribution of mean 0 and standard deviation 1 in the normal distribution)

2.4. Report the measurement

The report of the measurement can be expressed as

1) Flat load

Mf * (t) = Mf (t) + RanXucal, f + RannXusig, f

Mf (t) = 85.0 ucal, f = 3.810, usig, f = 1.703

If the confidence level is around 95%, the coverage factor is k = 2 in the normal distribution

Mf * (t) = 85.0 +/- 2 X 3.810 +/- 2 X 1.703.

The range of Mf * (t) is expressed as:

73.974 ≤ Mf * (t) ≤ 96.026 (confidence level approx. 95%, inclusion factor k = 2)

2) Edge load

Me * (t) = Me (t) + RanXucal, e + RannXusig, e

Me (t) = 700 ucal, e = 28.072, usig, f = 4.888

If the confidence level is around 95%, the coverage factor is k = 2 in the normal distribution

Me * (t) = 700.0 ± 2 X28.072 ± 2 X4.888.

The range of Me * (t) is expressed as follows.

634.080? Me * (t)? 765.920 (confidence level about 95%, inclusion factor k = 2)

2.5. Uncertainty overall

2.5.1. Smooth direction

2.5.2. Blade direction

Although the present invention has been described in connection with what is presently considered to be preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (7)

Samples of wind power generators;
A sensor unit for measuring a physical quantity of the sample; And
A control unit for deriving time series and statistical data from the measured physical quantity of the sensor unit and performing fatigue analysis and equivalent load analysis through a lane flow counting;
Including,
Wherein validation of the measured physical quantity at the sensor unit is implemented in at least one of the following data conditions,
Wherein the equivalent load is expressed by a value at a frequency of 1 Hz and is expressed by the following equation.
[Data condition]
Measurement error data generated by calibration of the sensor part, operation range of the sensor part, noise, etc., exclusion data of the measurement part of the sensor part and the data collection device outside the measurement range, reference data measured in a normal state, Data is measured in real-time mutual comparisons to verify suitability. Data measured under conditions such as wind speed distortions or wake effects caused by data, the terrain near the sample or other wind turbines or buildings, In the course of the yawing of the nacelle due to the change of the wind direction,
[Mathematical Expression]

(Where R _eq is the equivalent load, Ri is the i th load, ni is the number of the i th load, n _eq is the number of cycles equivalent, and m is the SN curve slope of the material)
The method according to claim 1,
Wherein the sensor unit comprises at least one of a strain gauge, a displacement meter, a load cell, a torque, an accelerometer, and a speedometer.
3. The method of claim 2,
Wherein the installation position of the sensor section satisfies the following conditions.
- location where high strain on unit load occurs, position that can avoid load inflow path, position where uniform stress is measured (that is, place where high stress / strain gradient is not taken , Position where stress concentration is not applied), position to ensure sufficient space, location where temperature can easily be determined, position on material with uniform load, position of material that can easily fix or attach measuring equipment.
delete delete delete delete

Images