阅读说明：本技术 一种在役风力机叶片结构损伤检测装置、系统及方法 (In-service wind turbine blade structure damage detection device, system and method ) 是由 童博 赵勇 高晨 王新 宋子琛 李立勋 陈臣 韩毅 于 2021-06-29 设计创作，主要内容包括：本发明提供一种在役风力机叶片结构损伤检测装置、系统及方法,包括以下步骤：步骤1,采集无损伤缺陷状态下待检测区域的声压时域信号；步骤2,对步骤1中得到的声压时域信号进行处理,得到声压功率谱密度；步骤3,将步骤2中得到的声压功率谱密度进行划分得到若干个频带,并计算各个频带对应的平均功率；步骤4,根据步骤3中得到的各个频带对应的平均功率,计算所有平均功率对应的判定区间；步骤5,根据步骤4中得到的判定区间,确定实测的风力机叶片是否存在叶片结构损伤；相对于传统的人工检测方法和安装传感器的长期检测方法,本发明简化了安装过程,降低了运维难度和风险,提高了在役风力机叶片结构损伤检测的可靠性和经济性。(The invention provides a device, a system and a method for detecting structural damage of an in-service wind turbine blade, which comprises the following steps: step 1, collecting sound pressure time domain signals of a region to be detected in a non-damage defect state; step 2, processing the sound pressure time domain signal obtained in the step 1 to obtain sound pressure power spectral density; step 3, dividing the sound pressure power spectral density obtained in the step 2 to obtain a plurality of frequency bands, and calculating the average power corresponding to each frequency band; step 4, calculating a judgment interval corresponding to all the average powers according to the average powers corresponding to the frequency bands obtained in the step 3; step 5, determining whether the actually measured wind turbine blade has blade structure damage according to the judgment interval obtained in the step 4; compared with the traditional manual detection method and the long-term detection method for installing the sensor, the method has the advantages that the installation process is simplified, the operation and maintenance difficulty and risk are reduced, and the reliability and the economy of the damage detection of the blade structure of the wind turbine in service are improved.)

1. A method for detecting structural damage of an in-service wind turbine blade is characterized by comprising the following steps:

step 1, collecting sound pressure time domain signals of a region to be detected in a non-damage defect state;

step 2, processing the sound pressure time domain signal obtained in the step 1 to obtain sound pressure power spectral density;

step 3, dividing the sound pressure power spectral density obtained in the step 2 to obtain a plurality of frequency bands, and calculating the average power corresponding to each frequency band;

step 4, calculating a judgment interval corresponding to all the average powers according to the average powers corresponding to the frequency bands obtained in the step 3;

and 5, determining whether the actually measured wind turbine blade has blade structure damage according to the judgment interval obtained in the step 4.

2. The in-service wind turbine blade structure damage detection method according to claim 1, wherein in step 2, the sound pressure time domain signal obtained in step 1 is processed to obtain a sound pressure power spectral density, and the specific method is as follows:

firstly, filtering the sound pressure time domain signal obtained in the step (1) to obtain a filtered sound pressure time domain signal;

secondly, carrying out fast Fourier transform on the obtained filtered sound pressure time domain signal to obtain sound pressure power spectral density.

3. The in-service wind turbine blade structural damage detection method according to claim 1, wherein in step 3, the sound pressure power spectral density obtained in step 2 is divided at intervals of S500HZ to obtain a plurality of frequency bands.

4. The in-service wind turbine blade structure damage detection method according to claim 1, wherein in step 3, the average power corresponding to each frequency band is calculated by the following formula:

wherein k is a frequency band number, and k is 1,2,3.. m; m is the number of divided frequency bands;is the average power of the k-th frequency band sound pressure signal.

5. The in-service wind turbine blade structure damage detection method according to claim 1, wherein in step 4, the determination intervals corresponding to all average powers are calculated according to the average powers corresponding to the frequency bands obtained in step 3, and the specific method is as follows:

firstly, respectively calculating the mean value and the standard deviation of the average power corresponding to each frequency band according to the average power corresponding to each frequency band obtained in the step 3;

and secondly, calculating to obtain a judgment interval corresponding to all average powers by combining a Lauda criterion according to the calculated mean value and standard deviation of the average powers corresponding to the frequency bands.

6. The in-service wind turbine blade structural damage detection method according to claim 5, wherein the determination intervals corresponding to all average powers are calculated by the following formula:

where Δ i is a determination interval.

7. The method for detecting structural damage of blades of an in-service wind turbine as claimed in claim 1, wherein in step 5, according to the determination interval obtained in step 4, it is determined whether the blades of the in-service wind turbine have structural damage, and the specific method is as follows:

collecting a sound pressure time domain signal of a wind turbine blade to be detected;

repeating the step 2 and the step 3 to obtain the average power corresponding to each frequency band of the wind turbine blade to be tested;

calculating the average value corresponding to all the average powers according to the average power corresponding to each frequency band of the wind turbine blade to be detected;

judging whether the mean value is in the judgment interval obtained in the step 4, wherein if the mean value is in the judgment interval obtained in the step 4, the actually measured wind turbine blade has blade structural damage; otherwise, the actual measurement wind turbine blade is not damaged.

8. A wind turbine blade damage detection system, characterized in that the system is capable of operating the method according to any one of claims 1 to 7, and comprises a sound pressure time domain signal acquisition unit, a sound pressure power spectral density acquisition unit, a division unit, a decision section acquisition unit and a damage level determination unit, in particular:

the sound pressure time domain signal acquisition unit is used for acquiring a sound pressure time domain signal of the area to be detected under a non-damage defect state;

the sound pressure power spectral density acquisition unit is used for processing the obtained sound pressure time domain signal to obtain sound pressure power spectral density;

the dividing unit is used for dividing the obtained sound pressure power spectral density to obtain a plurality of frequency bands and calculating the average power corresponding to each frequency band;

a judgment interval obtaining unit, configured to calculate a judgment interval corresponding to all average powers according to the obtained average powers corresponding to the frequency bands;

and the damage grade determining unit is used for determining whether the actually measured wind turbine blade has blade structure damage according to the obtained judgment interval.

9. The wind turbine blade damage detection device is characterized by comprising an acoustic exciter, an acoustic receiving unit and a data processing unit, wherein the acoustic exciter is used for generating an excitation sound source; the sound receiving unit is used for receiving sound waves generated after the sound source and the blade structure act, converting the received sound waves into detection data and then transmitting the detection data to the data processing unit; and the data processing unit is used for judging whether the actually measured wind turbine blade has blade structure damage according to the received detection data.

Technical Field

The invention belongs to the field of in-service wind turbine blade structure damage detection, and relates to an in-service wind turbine blade structure damage detection device, system and method.

Background

The service environment of the wind turbine blade is severe, the aging and structural damage of the blade material can be caused by wind load, sand scouring, rain erosion, salt spray corrosion, lightning stroke, ice coating and other factors in the operation process, if the factors are not found and treated in time, the damage expansion can cause the blade to break, catastrophic accidents are caused, and the operation safety of a unit is seriously threatened. The detection of the structural damage of the blades of the wind turbine in service is beneficial to timely finding out the abnormity of the operation of the blades, so that preventive maintenance is timely carried out, and the operation safety of a unit is improved.

At present, the existing in-service wind turbine blade structure damage detection technology comprises the steps that manual inspection is carried out on the outer surface of a blade through a hanging basket mode by detection personnel, optical and infrared inspection is carried out on the outer surface of the blade through an unmanned aerial vehicle, the detection personnel enter the interior of the blade to carry out manual internal inspection, and vibration and a strain sensor are installed in the blade to carry out long-term detection. The external detection risk of the blade is large, the accuracy is low, and the internal structural fault cannot be accurately found. The inside testing personnel of blade can only get into the position that the blade is less than half, and the detection range is not comprehensive enough, lacks scientific detection means, need will detect the blade adjustment to horizontal position and lock during the operation simultaneously, and the risk is big, and the operating efficiency is low. The sensor is high in long-term detection cost, and once a problem occurs, the maintenance and the replacement are difficult, so that the popularization and the application of the in-service wind turbine blade structure damage detection method are seriously restricted. For blades of in-service wind turbines, a detection means which is easy to implement, high in economy and high in reliability is urgently needed to accurately identify blade damage.

Disclosure of Invention

The invention aims to provide a device, a system and a method for detecting structural damage of a blade of an in-service wind turbine, which are used for solving the defects of high operation risk, high installation and maintenance difficulty and high cost in the existing detection and detection technology.

In order to achieve the purpose, the invention adopts the technical scheme that:

the invention provides a method for detecting structural damage of an in-service wind turbine blade, which comprises the following steps:

step 1, collecting sound pressure time domain signals of a region to be detected in a non-damage defect state;

step 2, processing the sound pressure time domain signal obtained in the step 1 to obtain sound pressure power spectral density;

step 3, dividing the sound pressure power spectral density obtained in the step 2 to obtain a plurality of frequency bands, and calculating the average power corresponding to each frequency band;

step 4, calculating a judgment interval corresponding to all the average powers according to the average powers corresponding to the frequency bands obtained in the step 3;

and 5, determining whether the actually measured wind turbine blade has blade structure damage according to the judgment interval obtained in the step 4.

Preferably, in step 2, the sound pressure time domain signal obtained in step 1 is processed to obtain a sound pressure power spectral density, and the specific method is as follows:

firstly, filtering the sound pressure time domain signal obtained in the step (1) to obtain a filtered sound pressure time domain signal;

secondly, carrying out fast Fourier transform on the obtained filtered sound pressure time domain signal to obtain sound pressure power spectral density.

Preferably, in step 3, the sound pressure power spectral density obtained in step 2 is divided at intervals of S500HZ to obtain a plurality of frequency bands.

Preferably, in step 3, the average power corresponding to each frequency band is calculated by the following formula:

wherein k is a frequency band number, and k is 1,2,3.. m; m is the number of divided frequency bands;is the average power of the k-th frequency band sound pressure signal.

Preferably, in step 4, the determination section corresponding to all the average powers is calculated according to the average power corresponding to each frequency band obtained in step 3, and the specific method is as follows:

firstly, respectively calculating the mean value and the standard deviation of the average power corresponding to each frequency band according to the average power corresponding to each frequency band obtained in the step 3;

and secondly, calculating to obtain a judgment interval corresponding to all average powers by combining a Lauda criterion according to the calculated mean value and standard deviation of the average powers corresponding to the frequency bands.

Preferably, the decision intervals corresponding to all average powers are calculated by:

where Δ i is a determination interval.

Preferably, in step 5, it is determined whether the measured wind turbine blade has a blade structure damage according to the determination section obtained in step 4, and the specific method is as follows:

collecting a sound pressure time domain signal of a wind turbine blade to be detected;

repeating the step 2 and the step 3 to obtain the average power corresponding to each frequency band of the wind turbine blade to be tested;

calculating the average value corresponding to all the average powers according to the average power corresponding to each frequency band of the wind turbine blade to be detected;

judging whether the mean value is in the judgment interval obtained in the step 4, wherein if the mean value is in the judgment interval obtained in the step 4, the actually measured wind turbine blade has blade structural damage; otherwise, the actual measurement wind turbine blade is not damaged.

A wind turbine blade damage detection system can operate the method and comprises a sound pressure time domain signal acquisition unit, a sound pressure power spectral density acquisition unit, a dividing unit, a judgment interval acquisition unit and a damage grade determination unit, and specifically comprises the following steps:

the sound pressure time domain signal acquisition unit is used for acquiring a sound pressure time domain signal of the area to be detected under a non-damage defect state;

the sound pressure power spectral density acquisition unit is used for processing the obtained sound pressure time domain signal to obtain sound pressure power spectral density;

the dividing unit is used for dividing the obtained sound pressure power spectral density to obtain a plurality of frequency bands and calculating the average power corresponding to each frequency band;

a judgment interval obtaining unit, configured to calculate a judgment interval corresponding to all average powers according to the obtained average powers corresponding to the frequency bands;

and the damage grade determining unit is used for determining whether the actually measured wind turbine blade has blade structure damage according to the obtained judgment interval.

A wind turbine blade damage detection device comprises an acoustic exciter, an acoustic receiving unit and a data processing unit, wherein the acoustic exciter is used for generating an excitation sound source; the sound receiving unit is used for receiving sound waves generated after the sound source and the blade structure act, converting the received sound waves into detection data and then transmitting the detection data to the data processing unit; and the data processing unit is used for judging whether the actually measured wind turbine blade has blade structure damage according to the received detection data.

Compared with the prior art, the invention has the beneficial effects that:

the invention provides a wind turbine blade damage detection method, which comprises the steps of actively transmitting and exciting a noise aid and receiving sound waves generated after a sound source and a blade structure act, calculating to obtain the average power of the sound waves after the sound waves act on the blade structure, deploying the system in a manufacturing plant to obtain the average power characteristic under the normal condition of a structurally-faultless blade, and then deploying the system in the same type of blade of a wind turbine generator to detect, wherein when the blade has a structural fault, the sound waves are transmitted to the outside of the blade through a penetrating structural fault, so that the loss of sound energy and the reduction of the average power of the sound waves received by the system are caused.

The invention provides a wind turbine blade damage detection device which comprises a sound exciter, a sound receiving unit and a data processing unit, wherein the sound exciter, the sound receiving unit and the data processing unit are deployed inside blades of a wind turbine generator in service during detection, the average power of the sound waves after the sound waves are acted with the blade structures is calculated by actively transmitting and exciting the sound aid and receiving the sound waves after the sound source is acted with the blade structures, and the detection data and normal data are compared to judge whether structural defects exist in the blades. Compared with the traditional manual detection method and the long-term detection method for installing the sensor, the method has the advantages that the installation process is simplified, the operation and maintenance difficulty and risk are reduced, and the reliability and the economy of the damage detection of the blade structure of the wind turbine in service are improved.

Drawings

FIG. 1 is a flow chart relating to the present invention;

fig. 2 is a schematic view of the mounting structure.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings.

As shown in fig. 1 and 2, in order to overcome the problems of low reliability, high operation risk and poor economy of the existing wind turbine blade structure damage detection technology, the invention provides a wind turbine blade structure damage detection system and method based on the active acoustic technology.

The detection system comprises: the device comprises an acoustic exciter, an acoustic receiving unit and a data processing unit; the acoustic exciter is used for generating an exciting sound source; the sound receiving unit is used for receiving sound waves generated after the sound source and the blade structure act, and comprises a microphone and an audio conversion module, wherein the audio conversion device is connected with the storage module and is used for converting sound wave signals into detection data; the sound receiving unit is provided with a data interface and is used for transmitting the received audio detection data to the data processing unit. The data processing unit is used for storing and analyzing the sound wave data.

Specifically, the method comprises the following steps:

step 1, deploying a detection system on a blade without structural defects.

Selecting a defect-free blade installation detection system in a blade manufacturing plant, and deploying an acoustic exciter, an acoustic receiving unit and a data processing unit in a partition plate positioned at the root of a wind turbine blade;

and 2, collecting sound pressure signals.

And (3) exciting an acoustic signal by adopting an acoustic excitation device, acquiring a sound pressure time domain signal of the blade in a defect-free state by utilizing an acoustic receiving unit, recording the signal as P and storing.

And 3, processing and analyzing the sound pressure signal.

Firstly, filtering the sound pressure time domain signal P obtained in the step (2) by adopting a high-pass filter to obtain a filtered sound pressure time domain signal L;

performing fast Fourier transform on the sound pressure time domain signal L after wave generation respectively according to the following formula to obtain sound pressure power spectral density S;

f is the frequency; sf is sound pressure power spectrum density corresponding to the frequency f; Δ t is the window time interval; n is the total number of sound pressure signal data; n is the number of sound pressure data in the window; h (n) is a window function; p (n) is the sound pressure time domain signal in the window; e is natural index base; j is an imaginary unit; and pi is the circumferential ratio.

Dividing sound pressure power spectral density at intervals of S500HZ to obtain a plurality of frequency bands; calculating the average power corresponding to the k-th frequency band by:

wherein k is a frequency band number, and k is 1,2,3.. m; m is the number of divided frequency bands;is the average power of the k-th frequency band sound pressure signal.

And 4, determining the judgment interval.

Average power in each frequency band obtained in step 3Performing statistical analysis to calculate all average powersThe corresponding mean value:

wherein m is the number of frequency bands,for all average powersThe corresponding mean value.

Calculating all average powers l^(k)The corresponding standard deviation:

where std is the average power of all^(k)The corresponding standard deviation.

Calculating the judgment intervals corresponding to all average powers according to the Lauda criterion by using the calculated mean value and standard deviation:

where Δ i is a determination interval.

And 5, identifying the damage.

According to the sensor layout scheme in the step 1, a detection system is installed inside blades of the same type of the in-service wind turbine generator to be detected, actual measurement sound pressure signals of the blades are collected according to the step 2, the actual measurement signals are processed according to the steps 3 to 4, and the mean value of average power corresponding to the actual measurement sound pressure signals is obtainedAccording toJudging whether the blade structure of the actually measured wind turbine blade is damaged or not, wherein if the mean value is in the judgment interval obtained in the step 4, the blade structure of the actually measured wind turbine blade is damaged; otherwise, the actual measurement wind turbine blade is not damaged.

Example of the implementation

Step one, a detection system is deployed on a blade without structural defects.

The in-factory structural defect-free fan blade selected in this example was 46.7m in length. The Yamaha CBR10 type loudspeaker is used as an acoustic excitation device, the PCB acoustic sensor 378A21 is used as an acoustic receiving unit, the excitation device and the receiving device are respectively arranged at the root part of the wind turbine blade, the loudspeaker excites an acoustic source, and the acoustic sensor receives acoustic waves after the structure/damage of the blade.

And step two, collecting sound pressure signals.

And opening a loudspeaker to excite an acoustic signal, and acquiring and storing a blade sound pressure time domain signal by using an acoustic receiving sensor. The sampling frequency was 51200Hz and the sampling time was 60 s.

And step three, processing and analyzing the sound pressure signal.

And filtering the sound pressure signal obtained in the step three by using a high-pass filter, wherein in the example, a Butterworth filter is selected for high-pass filtering, the cut-off frequency is 500Hz, and the frequency range of the sound pressure signal L obtained after filtering is 500 Hz-20000 Hz. And respectively carrying out fast Fourier transform on the L signals to obtain sound pressure power spectrum density S:

wherein: f is frequency, sf is sound pressure power spectrum density corresponding to f frequency, Δ t is window time interval, in this example, Δ t is 0.08S, N is total number of sound pressure signal data, in this example, N is 3072000, N is number of sound pressure data in window, in this example, N is 4096, h (N) is window function, in this example, rectangular window is used as window function, p (N) is sound pressure time domain signal in window, e is natural index base, j is imaginary unit, and pi is circumferential ratio.

Dividing according to the frequency bandwidth of 500Hz, and calculating the average power in each frequency band:

wherein k is a frequency band number, and k is 1,2,3.. m; m is the number of divided bands, in this example, m is 39; l^(k)Is the average power of the k-th frequency band sound pressure signal.

And step four, determining the interval.

Average power l of each frequency band of the processed sound pressure signal obtained in the step five^(k)Performing statistical analysis to calculate average power l^(k)Average value of (d):

wherein the content of the first and second substances,for each frequency band l^(k)Is measured.

Calculating the average power l^(k)Standard deviation:

wherein std is within each frequency bandStandard deviation of (2).

And calculating the judgment interval of the average power according to the Lauda criterion by using the mean value and the standard deviation of the average power under the condition of no structural damage.

Where Δ i is a determination interval.

Mean value of average power obtained in this exampleStandard deviation std_iAnd a damage determination section.

std_i＝6.24，Δi＝(121.41,158.85)

And identifying five kinds of damage.

According to the sensor layout scheme in the step one;

acquiring an actually measured sound pressure signal of the structural defect blade according to the step two;

processing the measured signal according to the steps from three to four;

finally, the average power value of each frequency band of the actual measurement sound pressure signal of the blade to be detected is obtained asAccording toAnd judging the damage of the leaves. The result shows that the method can effectively detect the damage of the wind turbine blade.