CN109209783A - A kind of method and device of the lightning damage based on noise measuring blade - Google Patents
A kind of method and device of the lightning damage based on noise measuring blade Download PDFInfo
- Publication number
- CN109209783A CN109209783A CN201811086071.2A CN201811086071A CN109209783A CN 109209783 A CN109209783 A CN 109209783A CN 201811086071 A CN201811086071 A CN 201811086071A CN 109209783 A CN109209783 A CN 109209783A
- Authority
- CN
- China
- Prior art keywords
- blade
- noise
- voice data
- damage
- lightning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The present invention relates to a kind of methods of lightning damage based on noise measuring blade, including the following steps: a method of the lightning damage based on noise measuring blade, including the following steps: voice data is measured by sound transducer during blade operation；The lightening activity near blade is determined by voice data；The voice data is analyzed in the case where determining lightening activity to determine noise spectrum；And determine whether blade has damaged according to noise spectrum.Moreover, it relates to a kind of device of the lightning damage based on noise measuring blade.Through the invention, the lightning damage of blade can be detected in real time, especially because of damage caused by being struck by lightning, be enable to make subsequent remedial measure, such as automatic feedback control, manual intervention in time to reduce artificial detection cost and avoid further losing.
Description
Technical field
Present invention is generally directed to wind power generation fields, more specifically, are related to a kind of based on noise measuring blade
The method of lightning damage.Moreover, it relates to a kind of device of the lightning damage based on noise measuring blade.
Background technique
In recent years, clean energy resource field shows fast-developing trend.As the representative of clean energy resource, wind-driven generator
Using growing.The blade of wind-driven generator is the core component of wind-driven generator capture wind energy, and health status is direct
It is related to equipment safety and generating efficiency.
The damage check of wind power generating set (hereinafter referred to as blower) blade at present, mainly by telescope or hanging basket,
The artificial visual type such as spider-man is completed, such as the lightning damage of blade, main by manually visualizing, lightning protection system is led
The general character and its resistance measurement, the modes such as thunder and lightning registration card or sensor of thunder and lightning carry out.
However, there is these artificial blade detection methods periodically or non-periodically lag, at high cost, downtime length etc. to lack
Point.
Especially the lightning damage detection of blade can only shut down and be checked after lightning stroke occurs.The inspection is usually all
It is the longer period of time after blade is struck by lightning, few then a couple of days, more then several months, progress, the hysteresis quality of inspection will lead to blade later
Run in spite of wound, this is likely to cause blade injury further expansion, thus make to repair it is more complicated, and it is serious when will lead to blade without
Method repairs and replaces and the major accidents such as leaf destruction even occur.
Currently, a kind of means for thunder and lightning detection are thunder and lightning registration cards, another kind is sensor of thunder and lightning.
But thunder and lightning registration card can only record peak point current, and cannot reflect the time that lightning current occurs, it can not be with control
System processed forms closed loop feedback.Such as registration card can only selective log history peak inrush current, be struck by lightning without recording
Number.
Sensor of thunder and lightning is able to record the time and number that thunder and lightning occurs, and closed loop feedback can be formed with control system, but
It is that sensor of thunder and lightning can only reflect that lightning current passes through, and cannot effectively differentiate lightning damage whether occurs, this still may cause
Unnecessary shutdown or additional inspection fee.
The Chinese patent application CN201810106452.6 of entitled " a kind of fan blade damage sync detection device " is disclosed
A kind of blade injury detection device of included infrared equipment, can more careful must check be checked from blade root to blade tip,
But the technology can not timely feedback the damage of blade, need when still falling within regular or non-periodically artificial passive detection, and checking
Longer downtime is wanted, the movement of inspection synchronous with fan operation can not carry out, and be related to generated energy loss.It is similar therewith, name
For the Chinese patent application of " wind electricity blade defect damage inspection method and inspection system based on unmanned helicopter "
CN201410407778.4 discloses a kind of strong using the unmanned plane progress blade for being equipped with the acquisition equipment such as infrared, sound, image
Health status monitoring.Current most of unmanned helicopter is both needed to artificially for equipment to be carried to target seat in the plane, shuts down and manually adjusts
Blade is that fixed pose just can be carried out automatic monitoring.The method still can not be realized really and timely be monitored on-line.
The Chinese patent application of entitled " a kind of wind power generating set fault diagnosis system and method "
CN201610077410.5 discloses a kind of method detection host part damage by sound transducer detection machine interior noise
Method, but this method is not related to blade injury and blade lightning damage, and the program is made an uproar due to such as cabin inner machine
The limitation of sound etc and can not be applied to blade injury detect.
The Chinese patent application of entitled " a kind of blade strain on-line measuring device of fiber grating "
CN201110447199.9 disclose it is a kind of using fibre strain monitor blade injury by the way of, which is a kind of preferable leaf
The mode of piece primary structure such as girder, rear beam damage, but for the lightning damage in tip region often occurs, hole of being such as struck by lightning,
The damage of aluminium blade tip, smaller area crack, which is often difficult to identify, and for in-service blade tip region due to operating space
Limitation cannot be introduced into internal progress Optical Fiber Strain Gauge arrangement substantially.Therefore the monitoring of lightning damage is lacked timely and accurately
Property.It is similar therewith, entitled " a kind of to monitor the method for the state of blade in wind power plant, calculate equipment and system " China
Patent application CN201710617375.6 is referred to the overall stiffness variation using acceleration transducer monitoring blade, to identify
The damage of primary structure.
The Chinese patent application of entitled " for estimating the system of the state by the non-conducting hollow structure being struck by lightning "
CN201110221622.3 provides a kind of lightning damage monitoring method based on pressure perception, and the method is not necessarily limited by for wind
The lightning damage of machine blade monitors.The method system complex needs installing monitoring device in blade inner cavity, need to additionally consider to set
It is standby itself to draw thunder and Lightning Protection；And this method depends on the susceptibility to pressure to the identification of lightning damage, as sensitivity is spent
The blade interior pressure change that high then normal blade movement generates may trigger wrong report, such as excessively insensitive, before not resulting in
Rear cracking rank damage be such as struck by lightning hole, lightning stroke crackle blade interior pressure change may because change it is unobvious due to can not be very
Good identification.
The Chinese patent application CN201710356740.2 of entitled " lightning stroke at positioning wind turbine " discloses one kind
The method of blade lightning stroke position based on lightning stroke sound and microphone array.This method is closed using the successive logic of thunder and lightning
System；Lightning stroke position is determined in conjunction with time difference of the microphone array to sound detection.This method does not refer to blade lightning stroke damage directly
Wound monitoring, can only monitor the approximate region be struck by lightning and its be struck by lightning.Due in blade operational process there are variable pitch, yaw and
The complex behaviors such as rotation speed change are difficult to accurately distinguish whether lightning stroke occurs arrester itself or its near zone.Therefore, it is difficult to
Effectively quickly identify blade whether by lightning damage and its severity.And accurately identifying for flash-point can be because of the change of intensity of sunshine
Change is a greater impact.
The Chinese patent application of entitled " a kind of running blade of wind-driven generator detection device of wind field "
CN201520586958.3 discloses one kind by establishing a height bracket appropriate near blower, and there are multiple points for bracket
Branch, and several microphone arrays are arranged in each branch.Blade is monitored on-line by sound, Image Acquisition.The party
Simultaneously the lightning damage for how monitoring and identifying blade is not specifically mentioned in method.
The Japan of entitled " the abnormal system of determination and the method for the abnormal system of determination of being struck by lightning to be fixed to blower of being struck by lightning "
Patent application JP2017181410, which is disclosed by the way of lightning current and vibration monitoring, identifies abnormal lightning stroke.The method will
Traditional lightning current monitoring is combined with vibration monitoring, but this method can not distinguish lightning stroke and lightning damage, when blade is normal
When connecing sudden strain of a muscle, it is also possible to triggering cause such as be struck by lightning hole, be struck by lightning crackle equivalent damage when same order vibration.
Summary of the invention
From the prior art, the task of the present invention is provide a kind of method of lightning damage based on noise measuring blade
And device can detect the damage that blade is subject to by lightning stroke by the method or device in real time after thunder and lightning occurs, thus
Make it possible to make subsequent remedial measure, such as automatic feedback control, manual intervention in time to reduce artificial detection cost
And it avoids further losing.
In the first aspect of the present invention, foregoing task by a kind of method of lightning damage based on noise measuring blade come
It solves, this method includes the following steps:
Voice data is measured by sound transducer during blade operation；
The lightening activity near blade is determined by voice data；
The voice data is analyzed in the case where determining lightening activity to determine noise spectrum；And
Determine whether blade has damaged according to noise spectrum.
It is provided in a preferred embodiment of the invention, determines that the lightening activity near blade includes: by voice data
The typical frequency spectrum mould of thunder in signal and/or identification voice data by being higher than thunder energy threshold value in identification voice data
Formula, to identify the lightening activity occurred near wind-driven generator.By the preferred embodiment, thunder and lightning can be identified at low cost
Activity.It should be pointed out that the mode of other detection lightening activities is also possible, such as identification acoustic pattern (thunder feature).
It provides in another preferred embodiment of the invention, this method further include:
Being associated between type of impairment and damage noise characteristic spectrum is established by machine learning.
By the preferred embodiment, being associated between various type of impairment and corresponding noise characteristic spectrum can be simply determined.
For example, noise spectrum has significant when lightning stroke cracking occurs, leading edge burn into blade tip falls off, drains hole plug equivalent damage type
The characteristics of, and machine learning can learn these features to identify for subsequent type of impairment.Damage noise characteristic music score such as
It may include the frequency spectrum and energy spectrum or intensity spectrum of noise.
It is provided in another preferred embodiment of the invention, determining whether blade has damaged according to noise spectrum includes:
Noise spectrum is compared with damage noise characteristic spectrum to determine similarity；And
Type of impairment associated with the damage noise characteristic spectrum is determined when similarity is higher than threshold value.
By the preferred embodiment, type of impairment can be simply determined.Due to the cracking that is such as struck by lightning, leading edge burn into blade tip
The many type of impairment for falling off, draining hole plug or the like are associated with unique blade noise, therefore with this solution can be with
Reliably identify various type of impairment.For example, front and rear edge unsticking, the profile variations such as lightning stroke hole, lightning stroke cracking or blade tip fall off or
Damage/defect will lead to local air flow changes in distribution, and the noise of high-frequency noise, higher-energy can be generated during blade movement
Or there is obvious noise variance in one or two of three blades, therefore can be by comparing voice data collected and damage
Noise characteristic is composed to reliably determine type of impairment.Noise characteristic spectrum is damaged either prespecified, machine can also be passed through
Device learns to obtain.This compares either by the noise level at single time point or frequency compared with single noise threshold
Compared with being also possible to the noise level for acquiring multiple time points or frequency and multiple noise threshold phases in stall noise characteristic spectrum
Compare, to more accurately determine type of impairment.
It provides in another preferred embodiment of the invention, damage noise characteristic spectrum includes: to determine that blade is according to noise spectrum
No damaged includes: the sound spectrum difference for identifying lower three blades of same rotational speed；And/or by the energy model or frequency of noise spectrum
Rate mode is compared with the history noise energy mode of the three of same Fans blades or typical spectrum mode.It is excellent by this
Scheme is selected, can determine damage at low cost.It should be noted here that in the present invention, energy refers to that acoustic energy, i.e. sound pressure exist
Integral in corresponding spectral range.In addition, machine learning can also be carried out by the preferred embodiment.
It is provided in an expansion scheme of the invention, the type of impairment includes: lightning stroke cracking, leading edge burn into blade tip
It falls off and drains hole plug.It should be noted here that the type of impairment not instead of exclusive list, in religion of the invention
It leads down, it can be envisaged that other type of impairment.
It is provided in another expansion scheme of the invention, voice data is measured by sound transducer during blade operation
Include:
Pass through the multiple sound numbers of the sound transducer array measurement being made of multiple sound transducers during blade operation
According to；
Wherein this method further include:
Damage position is determined by the multiple voice data.
By the preferred embodiment, accurate auditory localization may be implemented and be achieved in damage reason location.For example, passing through multiple wheats
Noise in the voice data of gram wind is strong and weak and combines the installation site of each microphone, and sound source can be determined relative to each wheat
The relative position of gram wind, so that it is determined that the position of sound source.
It is provided in a preferred embodiment of the invention, analyzing the voice data to determine noise spectrum includes:
Voice data is filtered to obtain the voice data of certain frequency range.
By the preferred embodiment, irrelevant signal can be filtered out.Such as it can be filtered out by high pass or bandpass filter
1000Hz machine noise below.Analysis voice data can also be mentioned including other technological means, such as signal enhancing, signal
It takes, pattern-recognition etc..
It provides in another preferred embodiment of the invention, this method further comprises the steps of:
The lightening activity near blade is determined by voice data.
With this solution, lightning damage can be effectively identified after lightening activity.For example, being identified by voice data
To after lightening activity, damage inspection can be carried out immediately.For example, can use typical higher than certain energy threshold A and its thunder
Frequency mode, identify and lightening activity near seat in the plane occur, then changed by sound spectrum and its three blade energy
It measures difference and determines that the logical relation with lightening activity occurs for damage.Such as the two small Mr. Yu's threshold interval time B of time of origin, i.e.,
It can be considered high probability lightning damage.Further combined with lightning stroke hole, the machine learning of the typical losses mode such as lightning stroke cracking is final to know
Other lightning damage, and alarm is provided, according to injury severity score feedback control system.
In the second aspect of the present invention, foregoing task by a kind of device of lightning damage based on noise measuring blade come
It solves, which includes:
One or more sound transducers are configured as measuring sound number by sound transducer during blade is run
According to；And
Controller is configured as executing following movement:
The lightening activity near blade is determined by voice data；
The voice data is analyzed in the case where determining lightening activity to determine noise spectrum；With
And
Determine whether blade has damaged according to noise spectrum.
It is provided in a preferred embodiment of the invention, one or more of sound transducers include being passed by multiple sound
The sound transducer array that sensor is constituted, and the controller is additionally configured to by being provided by the sound transducer array
Multiple voice datas determine damage position.
By the preferred embodiment, damage position can be reliably determined.
The present invention at least has following the utility model has the advantages that (1) is by the invention it is possible to be accurately determined lightning damage and damage
Hurt type, this is that the following uniqueness based on inventor is seen clearly: when lightning stroke causes blade injury, blade is probably due to such as open
Split, the lightning damage that hole of being struck by lightning, blade-section fall off or the like and generate larger profile variation, so that more significant make an uproar will be generated
Sound, thus by using microphone measure noise, can accurately identify whether occur lightning stroke and possible faulted condition, and by
Unique noise pattern will be presented in various damages, therefore can further determine that type of impairment by analyzing noise, thus in thunder
Electricity immediately determines that damage and type of impairment after occurring, and further determines that processing urgency level and processing scheme；(2) by this hair
It is bright, without carrying out larger transformation to blower itself, but by arranging one or more sound transducer, such as microphone and controls
Non-destructive tests can be realized in device, therefore the program can be highly suitable for existing blower；(3) through the invention, it can also send out
The blade of raw damage or blade position, this is because the present invention is made an uproar using the multiple microphones of arrangement at different locations to measure
Sound, so as to, come localization of sound source, thereby determine that damage by the position of measured multiple voice datas and corresponding microphone
Traumatic part position.
Detailed description of the invention
With reference to specific embodiment, the present invention is further explained with reference to the accompanying drawing.
Fig. 1 shows the schematic diagram of the device of the lightning damage according to the present invention based on noise measuring blade；And
Fig. 2 shows the flow charts of the method for the lightning damage according to the present invention based on noise measuring blade.
Specific embodiment
It should be pointed out that each component in each attached drawing may be shown in which be exaggerated in order to illustrate, and it is not necessarily ratio
Example is correctly.In the drawings, identical appended drawing reference is equipped with to the identical component of identical or function.
In the present invention, unless otherwise indicated, " on being arranged in ... ", " being arranged in ... top " and " on being arranged in ... "
Do not exclude the case where there are intermediaries therebetween.
In the present invention, each embodiment is intended only to illustrate the solution of the present invention, and is understood not to restrictive.
In the present invention, unless otherwise indicated, quantifier "one", " one " and the scene for not excluding element.
It is also noted herein that in an embodiment of the present invention, for it is clear, for the sake of simplicity, might show only one
Sub-unit or component, but those skilled in the art are it is understood that under the teachings of the present invention, it can be according to concrete scene
Need to add required component or component.
It is also noted herein that within the scope of the invention, the wording such as " identical ", " equal ", " being equal to " are not meant to
The two numerical value is absolutely equal, but allows certain reasonable error, that is to say, that the wording also contemplated " substantially phase
Together ", " being essentially equal ", " being substantially equal to ".
In addition, the number of the step of each method of the invention limit the method step execute sequence.Unless special
It does not point out, various method steps can be executed with different order.
Blower is normally operated in outside of the city, is far from the crowd, this causes some serious damages, such as blade tip to be lost, front and rear edge, abdomen
Plate cracking, girder perforation equivalent damage, cannot be found in time.These damages can be extended to bigger destructive knot in operation
Structure damage, and then lead to not repair and need integral replacing even leaf destruction.It by means of the present invention and system, can be with
Discovery damage in time, avoids damage to further expand, to effectively reduce shutdown and maintenance and replacement loss.
Especially in lightning stroke context of detection, the solution of the present invention, can not only be with control system shape compared with sensor of thunder and lightning
It at feedback closed loop, and can quickly and effectively determine lightning damage type, and feed back more detailed information, such as degree of injury, position
It sets and pattern information etc..
Compared with the existing technology, the present invention has following features: (1) the present inventor recognizes originally, due to the application
It is related to the blade lightning damage of concrete type, and the white noise that other host parts are formed in cabin can largely cover blade
Running noises, while cabin shell has great attenuation to blade sound in operation, therefore sensor arrangement is
Sound transducer position is optimized to outside cabin and tower, especially by optimal blade injury monitoring mode based on this
It is proximate to rotating vane and is most subject to the peaked area of lightning stroke be optimal position, is thus greatly improved detection accuracy；(2)
Using the distinctive sound pressure of thunder and spectrum signature the thunder identification near monitored blade is occurred for this patent, goes forward side by side one
Step combines the sound spectrum variation after blade injury, and the acoustic pressure order of magnitude of such as three blade mutual deviations, time shaft changes, pattern-recognition
Etc. modes, in conjunction with machine learning, thus monitoring blade lightning damage accurately and timely；(3) pass through Voice pattern recognition, the present invention
Lightning damage can be identified well, and avoids lightning stroke position in the prior art identification, Oscillation Amplitude identification etc. complexity side
Case.
Below by specific embodiment, the present invention is further explained.
Fig. 1 shows the schematic diagram of the device 100 of the lightning damage according to the present invention based on noise measuring blade.
As shown in Figure 1, device 100 include be arranged on wind-driven generator by multiple microphones 103,104,105,106
The microphone array of composition, in which:
Microphone 106 is arranged on pylon 110 at the position at the tip of blade 101 just to acquire blade 101
Noise at tip.It, can be preferable by the way that microphone 106 to be arranged on pylon 110 at the position at the tip of blade 101
Ground acquisition noise, because blade 101 is likely to vibrate and deform under the faulted condition for occurring such as to bend, be broken etc,
And it is often maximum to pass to the deformation at the tip of blade 101, therefore noise also can be clearest, therefore in its attachment cloth microphone
Can acquire clearest noise it is for processing and analysis use.
Microphone 103 and 104 is arranged in cabin 102 noise at the root to acquire blade 101 and machine is made an uproar
Sound.By acquiring machine noise, machine noise can be excluded in voice data collected.Microphone 103 is arranged in cabin
Sentencing for acquiring the environment noise such as machine noise far from wheel hub 108 on 102, and microphone 104 is arranged in cabin 102 and leans on
Noise at the root for sentencing acquisition blade 101 of nearly wheel hub 108.
Microphone 105 is arranged on pylon 110 at the position at the middle part of blade 101 to acquire in blade 101
Noise at portion.By a group microphone 103-106, the noise at the tip, middle part and root of blade can be comprehensively acquired, from
And it preferably determines sound source position and preferably determines whether blade damages and damage location.
Microphone 107, which can be optionally arranged at the root of pylon 110, to be installed and is safeguarded with aspect.In the embodiment
In, the unrelated noises such as the signal processing methods filtering environmental noises such as high-pass filter with suitable cutoff frequency can be passed through.
Device 100 further includes controller (not shown).Controller can both be arranged in wind-driven generator, such as cabin 102
It or in wheel hub 108, can also be arranged at separate wind-driven generator, be, for example, remote control computer.
Controller is configured as executing following movement:
The lightening activity near blade 101 is determined by voice data.Such as pass through one or more microphone 103-
106, voice data when lightning stroke occurs can be acquired, then analyze in voice data with the presence or absence of characteristic sounds frequency spectrum and/or
Greater than the intensity of sound of thunder energy threshold, in this case, then lightening activity has occurred in judgement.Lightening activity is occurring
In the case of, local form may be caused to change because being struck by lightning equivalent damage during blade movement, if leading edge is corroded, front and rear edge is de-
Viscous, lightning stroke hole, lightning stroke are cracked or profile variations or the damage/defects such as blade tip falls off.These profile variations or damage/defect can be led
Local air flow changes in distribution is caused, high-frequency noise, the noise of higher-energy or three leaves thus can be generated during blade movement
One of piece or two there is obvious noise variance.It, being capable of effectively automatic identification by the acquisition and analysis of sound transducer
Noise spectrum caused by the typical failures modes such as lightning stroke is cracked, leading edge burn into blade tip falls off, drains hole plug.Identify that thunder and lightning is living
Dynamic mode is, for example, and using certain energy threshold A and its typical frequency mode of thunder is higher than, identifies generation near seat in the plane
Lightening activity.After identifying lightening activity, damage hair is determined by sound spectrum variation and its three blade capacity volume variances
The raw logical relation with lightening activity.Such as the two small Mr. Yu's threshold time B of time of origin, that is, it can be considered high probability lightning stroke damage
Wound.Further combined with lightning stroke hole, the machine learning of the typical losses mode such as lightning stroke cracking finally identifies lightning damage, and provide
Alarm, according to injury severity score feedback control system.
The voice data is analyzed to determine noise spectrum.Such as controller can be to by one or more microphone 103-
One or more voice datas of 106 outputs are handled to extract interested voice signal as noise spectrum.The processing
Including filtering, signal amplification, pattern-recognition etc..Filtering may include being filtered out machine using bandpass filter or high-pass filter and being made an uproar
Sound, such as 1000Hz low-frequency noise below or other ambient noises.Further, it is also possible to analyze processed the multiple sound
Data are to determine sound source position and noise level.Determine the mode of sound source position for example are as follows: determine in each voice data
Noise type and noise intensity (such as using sound level meter)；Exclude non-stall noise and according to the noise intensity of each microphone
With the location estimation sound source position of corresponding microphone, for example determine the plane coordinates or sky of one or more sound sources in the spatial domain
Between coordinate.
Determine whether blade has damaged according to noise spectrum.Such as noise spectrum can be compared with damage noise characteristic spectrum
To determine similarity, and type of impairment associated with the damage noise characteristic spectrum is determined when similarity is higher than threshold value,
The threshold value can be empirically determined or be determined by machine learning.Since the cracking that is such as struck by lightning, leading edge burn into blade tip are de-
The many type of impairment for falling, draining hole plug or the like are associated with unique blade noise, therefore can reliably identify each
Kind type of impairment.For example, front and rear edge unsticking, the profile variations such as lightning stroke hole, lightning stroke cracking or blade tip fall off or damage/defect can be led
Local air flow changes in distribution is caused, high-frequency noise, the noise of higher-energy or three blades can be generated during blade movement
One or two there is obvious noise variance, therefore can come by comparing voice data collected and damage noise characteristic spectrum
Reliably determine type of impairment.Such as on blade 101 due to being struck by lightning there are in the case where significant crack 109, have and split
The noise of the blade 101 of seam 109 can have with other blades 101 with the noise of significant difference and crack generation significant special
Sign, therefore can determine that blade 101 is damaged and determines that corresponding type of impairment is crack 109 according to the difference.
Noise characteristic spectrum is damaged either prespecified, can also be obtained by machine learning.This, which compares, both may be used
To be that the noise level at single time point or frequency compare with single noise threshold, it is also possible to adopt multiple time points
The noise level or frequency of collection are compared with multiple noise thresholds in stall noise characteristic spectrum, to more accurately determine damage
Type.
Alerting signal is optionally provided in the case where determining that blade 101 damages, wherein alerting signal can be with
Including type of impairment and/or damage position.The alerting signal for example may include sound, light, haptic signal.
Being associated between type of impairment and damage noise characteristic spectrum is established alternately through machine learning.For example, can be with
The voice signal under various faulted conditions is recorded by one or more microphone 103-106, and by summarizing and analyzing, is determined
Being associated between the mode of these voice signals and corresponding type of impairment.Such as it for lightning damage, can be closed with thunder frequency
Connection, further control identification lightning stroke failure and its position, the type of failure etc..By machine learning, blade construction may recognize that
And the distinctive noise pattern of disfigurement, to significantly improve the timeliness and accuracy of judgement.Since machine learning algorithm is
It is known in the art, therefore in order not to the fuzzy present invention, elaboration is not further spread out herein.
Finally, device 100 also optionally includes actuator (not shown), such as electric motor.The actuator is configured as
The operating status of blade is adjusted according to alerting signal, such as faulted condition and damage position, for example, shuts down, slows down or adjust and attack
Angle etc. is to increase the safety in operation under degree of impairment.
The present invention at least has following the utility model has the advantages that (1) is by the invention it is possible to be accurately determined lightning damage and damage
Hurt type, this is that the following uniqueness based on inventor is seen clearly: when lightning stroke causes blade injury, blade is probably due to such as open
Split, the lightning damage that hole of being struck by lightning, blade-section fall off or the like and generate larger profile variation, so that more significant make an uproar will be generated
Sound, thus by using microphone measure noise, can accurately identify whether occur lightning stroke and possible faulted condition, and by
Unique noise pattern will be presented in various damages, therefore can further determine that type of impairment by analyzing noise, thus in thunder
Electricity immediately determines that damage and type of impairment after occurring, and further determines that processing urgency level and processing scheme；(2) by this hair
It is bright, without carrying out larger transformation to blower itself, but by arranging one or more sound transducer, such as microphone and controls
Non-destructive tests can be realized in device, therefore the program can be highly suitable for existing blower；(3) through the invention, it can also send out
The blade of raw damage or blade position, this is because the present invention is made an uproar using the multiple microphones of arrangement at different locations to measure
Sound, so as to, come localization of sound source, thereby determine that damage by the position of measured multiple voice datas and corresponding microphone
Traumatic part position.
Fig. 2 shows the processes of the method 200 of the lightning damage according to the present invention based on noise measuring blade, wherein empty
Wire frame representation optional step.
In optional step 202, being associated between type of impairment and damage noise characteristic spectrum is established by machine learning.The step
Suddenly it can be realized by known machine learning algorithm.
In step 204, voice data is measured by sound transducer during blade operation.The sound transducer is for example
It for microphone, and may include multiple microphones to acquire multiple voice datas.
In step 206, the lightening activity near blade is determined by voice data.After determining lightening activity, it can stand
Carry out damage check to determine the damage because of the noise that is struck by lightning.
In step 208, the voice data is analyzed to determine noise spectrum.The analytic process may include necessary signal processing
Process such as filters, amplifies, pattern-recognition.
In step 210, determine whether blade has damaged according to noise spectrum.Noise characteristic spectrum is damaged either prespecified
, it can also be obtained by machine learning.This compares either by the noise level at single time point or frequency and list
A noise threshold compares, and is also possible in the noise level for acquiring multiple time points or frequency and stall noise characteristic spectrum
Multiple noise thresholds compare, to more accurately determine type of impairment.
In optional step 212, alerting signal is provided in the case where determining that blade damages, wherein alerting signal may be used also
To include type of impairment and/or damage position.
In optional step 214, adjusted according to alerting signal, such as faulted condition and damage position blade operating status,
Such as shut down, slow down or adjust the angle of attack etc., to increase the safety in operation under degree of impairment.
Although some embodiments of the present invention are described in present specification, those skilled in the art
Member is it is understood that these embodiments are merely possible to shown in example.Those skilled in the art under the teachings of the present invention may be used
To expect numerous variant schemes, alternative solution and improvement project without beyond the scope of this invention.The appended claims purport
It is limiting the scope of the invention, and is covering the method in the range of these claims itself and its equivalents and knot whereby
Structure.
Claims (10)
1. a kind of method of the lightning damage based on noise measuring blade, including the following steps:
Voice data is measured by sound transducer during blade operation；
The lightening activity near blade is determined by voice data；
The voice data is analyzed in the case where determining lightening activity to determine noise spectrum；And
Determine whether blade has damaged according to noise spectrum.
2. according to the method described in claim 1, wherein determining that the lightening activity near blade includes: to pass through by voice data
Identify the typical spectrum mode of thunder in the signal for being higher than thunder energy threshold value in voice data and/or identification voice data,
To identify the lightening activity occurred near wind-driven generator.
3. according to the method described in claim 1, further include:
Being associated between type of impairment and damage noise characteristic spectrum is established by machine learning.
4. method according to claim 1 or 2, determining whether blade has damaged according to noise spectrum includes:
Identify the sound spectrum difference of lower three blades of same rotational speed；And/or
By the history noise energy mode or allusion quotation of the energy model of noise spectrum or three blades of frequency mode and same Fans
The spectrum mode of type compares.
5. according to the method described in claim 2, wherein determining whether blade has damaged according to noise spectrum and including:
Noise spectrum is compared with damage noise characteristic spectrum to determine similarity；And
Type of impairment associated with the damage noise characteristic spectrum is determined when similarity is higher than threshold value.
6. according to the method described in claim 3, wherein the type of impairment includes: that lightning stroke is cracked, leading edge burn into blade tip takes off
It falls, protect membrane damage, one or two zero vane scales knowledge dislocation and draining hole plug.
7. according to the method described in claim 1, wherein measuring voice data packet by sound transducer during blade operation
It includes:
The multiple voice datas of sound transducer array measurement during blade operation by being made of multiple sound transducers；
Wherein this method further include:
Damage position is determined by the multiple voice data.
8. according to the method described in claim 1, analyzing the voice data wherein to determine noise spectrum and including:
Voice data is filtered to obtain the voice data of certain frequency range.
9. a kind of device of the lightning damage based on noise measuring blade, comprising:
One or more sound transducers are configured as measuring voice data by sound transducer during blade is run；
And
Controller is configured as executing following movement:
The lightening activity near blade is determined by voice data；
The voice data is analyzed in the case where determining lightening activity to determine noise spectrum；And
Determine whether blade has damaged according to noise spectrum.
10. device according to claim 9, wherein one or more of sound transducers include by multiple sound sensors
The sound transducer array that device is constituted, and the controller is additionally configured to by being provided by the sound transducer array
Multiple voice datas determine damage position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811086071.2A CN109209783A (en) | 2018-09-18 | 2018-09-18 | A kind of method and device of the lightning damage based on noise measuring blade |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811086071.2A CN109209783A (en) | 2018-09-18 | 2018-09-18 | A kind of method and device of the lightning damage based on noise measuring blade |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109209783A true CN109209783A (en) | 2019-01-15 |
Family
ID=64984814
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811086071.2A Pending CN109209783A (en) | 2018-09-18 | 2018-09-18 | A kind of method and device of the lightning damage based on noise measuring blade |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109209783A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110364969A (en) * | 2019-07-05 | 2019-10-22 | 国电联合动力技术(连云港)有限公司 | A kind of restorative procedure of fan blade lightning protection system |
| CN110985310A (en) * | 2019-12-16 | 2020-04-10 | 大连赛听科技有限公司 | Wind driven generator blade fault monitoring method and device based on acoustic sensor array |
| CN111306010A (en) * | 2020-04-17 | 2020-06-19 | 北京天泽智云科技有限公司 | Method and system for detecting lightning damage of fan blade |
| CN111306008A (en) * | 2019-12-31 | 2020-06-19 | 远景智能国际私人投资有限公司 | Fan blade detection method, device, equipment and storage medium |
| CN111323115A (en) * | 2020-03-11 | 2020-06-23 | 宁波奥克斯电气股份有限公司 | Method and system for testing abnormal sound of indoor unit of air conditioner |
| CN112083077A (en) * | 2020-09-11 | 2020-12-15 | 上海萌郁智能科技有限公司 | Periodic motion blade audio fault detection system |
| CN112648135A (en) * | 2020-12-09 | 2021-04-13 | 华能浙江平湖海上风电有限责任公司 | Wind power blade and refitting process thereof |
| CN112727710A (en) * | 2020-12-15 | 2021-04-30 | 北京天泽智云科技有限公司 | Wind field thunderbolt density statistical method and system based on audio signals |
| CN113655340A (en) * | 2021-08-27 | 2021-11-16 | 国网湖南省电力有限公司 | Voiceprint recognition-based power transmission line lightning stroke fault positioning method, system and medium |
| WO2022063370A1 (en) * | 2020-09-23 | 2022-03-31 | Vestas Wind Systems A/S | Wind turbine monitoring device, wind turbine system, and wind turbine monitoring method |
| EP4030054A1 (en) * | 2021-01-18 | 2022-07-20 | Shanghai Baiantek Sensing Technology Co., Ltd. | Blade state monitoring method and device, storage medium, and wind power generator |
| EP4317680A1 (en) * | 2022-08-05 | 2024-02-07 | Christoph Lucks | Method for detecting an imbalance in a wind turbine and for generating electricity by means of a wind turbine |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040236538A1 (en) * | 1999-10-06 | 2004-11-25 | Aloys Wobben | System for monitoring wind power plants |
| US20060080418A1 (en) * | 2004-10-13 | 2006-04-13 | Wayne State University | Farfield analysis of noise sources |
| CN101937025A (en) * | 2009-06-29 | 2011-01-05 | 通用电气公司 | Be used to detect the system and method for thunder and lightning |
| US20110265575A1 (en) * | 2010-07-29 | 2011-11-03 | Glen Peter Koste | System for estimating a condition of non-conductive hollow structure exposed to a lightning strike |
| CN102828919A (en) * | 2011-06-14 | 2012-12-19 | 西门子公司 | Method for acoustically monitoring wind turbine, acoustic monitoring system and re-equipment kit |
| EP2565444A1 (en) * | 2011-08-31 | 2013-03-06 | Wölfel Beratende Ingenieure GmbH & Co. KG | Method and device for monitoring the status of rotor blades |
| CN104101652A (en) * | 2014-07-10 | 2014-10-15 | 南京航空航天大学 | Audio signal based wind power blade damage monitoring method and system |
| CN104677623A (en) * | 2015-03-16 | 2015-06-03 | 西安交通大学 | On-site acoustic diagnosis method and monitoring system for wind turbine blade failure |
| CN105136435A (en) * | 2015-07-15 | 2015-12-09 | 北京汉能华科技股份有限公司 | Wind generating set blade fault diagnosis method and device |
| EP3246563A1 (en) * | 2016-05-20 | 2017-11-22 | Siemens Aktiengesellschaft | Locating a lightning strike at a wind turbine |
| CN108138750A (en) * | 2015-10-07 | 2018-06-08 | 乌本产权有限公司 | For monitoring the method for wind energy plant |
-
2018
- 2018-09-18 CN CN201811086071.2A patent/CN109209783A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040236538A1 (en) * | 1999-10-06 | 2004-11-25 | Aloys Wobben | System for monitoring wind power plants |
| US20060080418A1 (en) * | 2004-10-13 | 2006-04-13 | Wayne State University | Farfield analysis of noise sources |
| CN101937025A (en) * | 2009-06-29 | 2011-01-05 | 通用电气公司 | Be used to detect the system and method for thunder and lightning |
| US20110265575A1 (en) * | 2010-07-29 | 2011-11-03 | Glen Peter Koste | System for estimating a condition of non-conductive hollow structure exposed to a lightning strike |
| CN102828919A (en) * | 2011-06-14 | 2012-12-19 | 西门子公司 | Method for acoustically monitoring wind turbine, acoustic monitoring system and re-equipment kit |
| EP2565444A1 (en) * | 2011-08-31 | 2013-03-06 | Wölfel Beratende Ingenieure GmbH & Co. KG | Method and device for monitoring the status of rotor blades |
| CN104101652A (en) * | 2014-07-10 | 2014-10-15 | 南京航空航天大学 | Audio signal based wind power blade damage monitoring method and system |
| CN104677623A (en) * | 2015-03-16 | 2015-06-03 | 西安交通大学 | On-site acoustic diagnosis method and monitoring system for wind turbine blade failure |
| CN105136435A (en) * | 2015-07-15 | 2015-12-09 | 北京汉能华科技股份有限公司 | Wind generating set blade fault diagnosis method and device |
| CN108138750A (en) * | 2015-10-07 | 2018-06-08 | 乌本产权有限公司 | For monitoring the method for wind energy plant |
| EP3246563A1 (en) * | 2016-05-20 | 2017-11-22 | Siemens Aktiengesellschaft | Locating a lightning strike at a wind turbine |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110364969A (en) * | 2019-07-05 | 2019-10-22 | 国电联合动力技术(连云港)有限公司 | A kind of restorative procedure of fan blade lightning protection system |
| CN110364969B (en) * | 2019-07-05 | 2020-09-22 | 国电联合动力技术(连云港)有限公司 | Repair method for lightning protection system of fan blade |
| CN110985310B (en) * | 2019-12-16 | 2021-03-30 | 大连赛听科技有限公司 | Wind driven generator blade fault monitoring method and device based on acoustic sensor array |
| CN110985310A (en) * | 2019-12-16 | 2020-04-10 | 大连赛听科技有限公司 | Wind driven generator blade fault monitoring method and device based on acoustic sensor array |
| CN111306008A (en) * | 2019-12-31 | 2020-06-19 | 远景智能国际私人投资有限公司 | Fan blade detection method, device, equipment and storage medium |
| CN111306008B (en) * | 2019-12-31 | 2022-03-11 | 远景智能国际私人投资有限公司 | Fan blade detection method, device, equipment and storage medium |
| CN111323115B (en) * | 2020-03-11 | 2022-06-14 | 宁波奥克斯电气股份有限公司 | Method and system for testing abnormal sound of indoor unit of air conditioner |
| CN111323115A (en) * | 2020-03-11 | 2020-06-23 | 宁波奥克斯电气股份有限公司 | Method and system for testing abnormal sound of indoor unit of air conditioner |
| CN111306010A (en) * | 2020-04-17 | 2020-06-19 | 北京天泽智云科技有限公司 | Method and system for detecting lightning damage of fan blade |
| CN111306010B (en) * | 2020-04-17 | 2022-03-29 | 北京天泽智云科技有限公司 | Method and system for detecting lightning damage of fan blade |
| CN112083077A (en) * | 2020-09-11 | 2020-12-15 | 上海萌郁智能科技有限公司 | Periodic motion blade audio fault detection system |
| WO2022063370A1 (en) * | 2020-09-23 | 2022-03-31 | Vestas Wind Systems A/S | Wind turbine monitoring device, wind turbine system, and wind turbine monitoring method |
| CN112648135A (en) * | 2020-12-09 | 2021-04-13 | 华能浙江平湖海上风电有限责任公司 | Wind power blade and refitting process thereof |
| CN112648135B (en) * | 2020-12-09 | 2022-01-11 | 华能浙江平湖海上风电有限责任公司 | Wind power blade and refitting process thereof |
| CN112727710A (en) * | 2020-12-15 | 2021-04-30 | 北京天泽智云科技有限公司 | Wind field thunderbolt density statistical method and system based on audio signals |
| EP4030054A1 (en) * | 2021-01-18 | 2022-07-20 | Shanghai Baiantek Sensing Technology Co., Ltd. | Blade state monitoring method and device, storage medium, and wind power generator |
| CN114810504A (en) * | 2021-01-18 | 2022-07-29 | 上海拜安传感技术有限公司 | Blade state monitoring method and device, storage medium and wind driven generator |
| CN113655340A (en) * | 2021-08-27 | 2021-11-16 | 国网湖南省电力有限公司 | Voiceprint recognition-based power transmission line lightning stroke fault positioning method, system and medium |
| CN113655340B (en) * | 2021-08-27 | 2023-08-15 | 国网湖南省电力有限公司 | Transmission line lightning fault positioning method, system and medium based on voiceprint recognition |
| EP4317680A1 (en) * | 2022-08-05 | 2024-02-07 | Christoph Lucks | Method for detecting an imbalance in a wind turbine and for generating electricity by means of a wind turbine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109209783A (en) | A kind of method and device of the lightning damage based on noise measuring blade | |
| JP7199608B2 (en) | Methods and apparatus for inspecting wind turbine blades, and equipment and storage media therefor | |
| CN101592590B (en) | Fault indirect diagnosis technique of rotating blade | |
| CN112067701B (en) | Fan blade remote auscultation method based on acoustic diagnosis | |
| US9080575B2 (en) | Method of detecting and controlling stall in an axial fan | |
| CN104075795A (en) | Method and system for monitoring vibration state of impeller of wind generating set | |
| CA2999829C (en) | Method for monitoring a wind turbine | |
| EP2660582B1 (en) | System and Method for Monitoring the Health of Stator Vanes | |
| CN102494894A (en) | Audio monitoring and fault diagnosis system for wind generating set and audio monitoring and fault diagnosis method for same | |
| CN104677623A (en) | On-site acoustic diagnosis method and monitoring system for wind turbine blade failure | |
| CN109715936A (en) | For monitoring the method and apparatus and computer program product of the state of at least one wind turbine | |
| CN103998775A (en) | Method for determining mechanical damage to a rotor blade of a wind turbine | |
| CN112324629A (en) | Wind power blade early damage monitoring system and method based on vibration and sound | |
| US8919202B2 (en) | System and method for monitoring health of stator vanes | |
| CN106706241B (en) | Active self-checking device and method for damage of wind turbine blade | |
| CN114450483A (en) | Detection of abnormal conditions of wind turbine generator | |
| CN109139390B (en) | Fan blade fault identification method based on acoustic signal feature library | |
| CN115640503B (en) | Wind turbine generator blade abnormality detection method and device | |
| Papasalouros et al. | Acoustic Emission Monitoring of Composite Blade of NM48/750 NEG-MICON Wind Turbine. | |
| US20230366382A1 (en) | Abnormality determination method for wind power generation device, abnormality determination system for wind power generation device, and abnormality determination program for wind power generation device | |
| CN109026555B (en) | Method and apparatus for identifying stall of blade | |
| CN111173687B (en) | On-line monitoring device and method for crack damage of wind power fan blade | |
| CN113504302B (en) | Method and system for monitoring fan blade state, electronic equipment and storage medium | |
| WO2022142213A1 (en) | Blade fault diagnosis method, apparatus and system, and storage medium | |
| CN111289179B (en) | Method for detecting unbalanced fusion of impellers of wind generating set |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190115 |
|
| RJ01 | Rejection of invention patent application after publication |