A wind turbine blade early damage monitoring system and method based on vibration and sound

Through a vibration and sound-based monitoring system, the problem of precise positioning difficulties and inability to achieve online monitoring in wind turbine blade monitoring is solved, efficient blade status analysis and fault detection are achieved, and operating costs and accident probability are reduced.

CN112324629BActive Publication Date: 2024-12-03HUANENG SHAANXI JINGBIAN ELECTRIC POWER CO LTD +2
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202011240487.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-11-09
Publication Date
2024-12-03
Estimated Expiration
2040-11-09

AI Technical Summary

Technical Problem

The prior art has problems in the monitoring of wind turbine blades, which are difficult to accurately locate the fault type, increase the risk of lightning, inability to realize online monitoring, shutdown inspection, resulting in power generation loss and low efficiency of operation and maintenance personnel.

Method used

Using an early damage monitoring system based on vibration and sound, the vibration signals and sound signals of the blade are extracted through acquisition, noise reduction and feature, and abnormal whistle profiles and vibration frequency are obtained, and the monitoring and analysis are carried out to realize contactless online monitoring to reduce the smallest The probability of damage evolve into a major accident reduces operating costs.

Benefits of technology

It realizes efficient online monitoring of wind turbine blades, reduces operating costs, increases the number of units available hours, reduces production and operation costs, and avoids economic losses caused by maintenance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112324629B_ABST
    Figure CN112324629B_ABST
Patent Text Reader

Abstract

The present invention discloses a system and method for monitoring early damage of wind turbine blades based on vibration and sound, belonging to the technical field of wind turbine status monitoring. First, the vibration signal and sound signal of the blade are collected, and then the collected vibration signal and sound signal are subjected to noise reduction processing; feature extraction is performed on the vibration signal after noise reduction to obtain the abnormal whistle profile of the blade; feature extraction is performed on the sound signal after noise reduction to obtain the vibration frequency and vibration mode of the blade; according to the abnormal whistle profile, vibration frequency and vibration mode of the blade, the status of the wind turbine blade is monitored and analyzed. The present invention has the characteristics of high detection efficiency, non-contact, low installation and operation cost, etc. through dual detection of vibration and sound waves, which can reduce the probability of small damage gradually evolving into a major accident and reduce operating costs; at the same time, it does not damage the structure of the wind turbine and does not affect the normal operation of the wind turbine.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of wind turbine status monitoring, and in particular relates to a system and method for monitoring early damage of wind turbine blades based on vibration and sound. Background Art

[0002] Blades are key components of wind turbines that absorb wind energy. If their performance deteriorates during operation, it will not only cause damage to the blades themselves, but also affect the performance and safety reliability of the entire unit. Since the working environment of blades belongs to the outdoor natural environment, they are affected by the environment during operation, especially lightning, rain, snow, gusts, etc., which may cause damage to the blades and even the wind turbine, and thus develop into equipment failure or even safety accidents, causing significant economic losses to the wind farm.

[0003] At present, the commonly used methods for blade monitoring include status monitoring and diagnosis based on main control operation data, strain monitoring and blade inspection using drones. Among them, status monitoring and diagnosis monitoring based on main control operation data rely on higher quality algorithms, which cannot accurately locate the type of fault and it is difficult to eliminate the impact of other parts of the wind turbine. Strain monitoring increases the risk of lightning strikes for blades and is insensitive to damage to coatings, protective films and pneumatic accessories. When using drones for blade inspections, it is impossible to achieve online monitoring of the blades, and a large amount of power generation is lost during shutdown inspections. At the same time, drones are affected by problems such as battery life, which greatly reduces the work efficiency of operation and maintenance personnel during the actual inspection process. Summary of the invention

[0004] In order to solve the defects existing in the above-mentioned prior art, the purpose of the present invention is to provide a wind turbine blade early damage monitoring system and method based on vibration and sound, which has high detection efficiency, does not damage the wind turbine structure, does not affect the normal operation of the wind turbine, and has low installation and maintenance costs.

[0005] The present invention is achieved through the following technical solutions:

[0006] The present invention discloses a wind turbine blade early damage monitoring system based on vibration and sound, comprising the following steps:

[0007] Step 1: Collect the vibration signal and sound signal of the blade;

[0008] Step 2: Perform noise reduction processing on the vibration signal and sound signal collected in step 1;

[0009] Step 3: Extract features from the vibration signal after noise reduction in step 2 to obtain the abnormal whistle profile of the blade; extract features from the sound signal after noise reduction to obtain the vibration frequency and vibration mode of the blade;

[0010] Step 4: Based on the abnormal whistle profile, vibration frequency and vibration mode of the blade obtained in step 3, monitor and analyze the status of the wind turbine blade.

[0011] Preferably, in step 1, the vibration signal of the blade is collected from inside the blade; and the sound signal of the blade is collected from outside the tower.

[0012] Preferably, in step 2, the noise reduction process is to inject white noise into the collected vibration signal and sound signal using the EEMD method, and then perform empirical mode decomposition and signal reconstruction.

[0013] Preferably, in step 3, the vibration signal is calculated using Fourier analysis, and after obtaining the vibration frequency and vibration mode of the blade, a comparative analysis is performed with the initial state of the blade when it is newly installed, and the change in each order of modal stiffness is calculated.

[0014] Preferably, in step 3, the Born-Jordan distribution is used to perform time-frequency analysis on the sound signal to obtain the abnormal whistle profile of the blade.

[0015] Preferably, in step 4, the data analysis results of the wind turbine blade status are sent to a remote data analysis platform for diagnostic analysis. The remote data analysis platform includes a blade failure mode rule library, and the blade failure mode is obtained according to the blade failure mode rule library.

[0016] The present invention discloses a wind turbine blade early damage monitoring system based on vibration and sound, comprising:

[0017] A vibration signal acquisition module, a sound signal acquisition module, a noise reduction module and a processor; the vibration signal acquisition module and the sound signal acquisition module are respectively connected to the noise reduction module, and the noise reduction module is connected to the processor; the vibration signal acquisition module and the sound signal acquisition module are used to execute step 1 of the method described in any one of claims 1 to 6, the noise reduction module is used to execute step 2 of the method described in any one of claims 1 to 6, and the processor is used to execute steps 3 and 4 of the method described in any one of claims 1 to 6.

[0018] Preferably, the vibration signal acquisition module includes a plurality of vibration sensors and a vibration signal collector. The plurality of vibration sensors are arranged and installed in the wind turbine blade cover plate. The vibration sensors are connected to the vibration signal collector, and the vibration signal collector is connected to the noise reduction module.

[0019] Preferably, the sound signal collection module includes a sound sensor and a sound signal collector. The sound sensor is arranged outside the tower, the sound sensor is connected to the sound signal collector, and the sound signal collector is connected to the noise reduction module.

[0020] Further preferably, the sound sensor is a plurality of silicon microphones arranged in an array.

[0021] Compared with the prior art, the present invention has the following beneficial technical effects:

[0022] The present invention discloses a method for monitoring early damage of wind turbine blades based on vibration and sound. The method uses vibration signals and sound signals to realize online monitoring of wind turbine blades. After noise reduction processing of the signals, the features for blade fault analysis are extracted from the sound signals and vibration signals respectively, which can reduce the probability of small damage gradually evolving into a major accident, reduce operating costs, and increase the number of hours available for the unit, thereby achieving the purpose of reducing production and operation costs. When the fan blades are operating under normal conditions, there is only the sound of the impeller sweeping the wind. When a fault occurs, abnormal sounds may occur, and the sound characteristics will change. In addition, the abnormal sounds will change accordingly under different operating conditions and health conditions. Therefore, according to the acoustic characteristics of the fan blades, the sound signal detection method and the corresponding algorithm can be used to effectively judge the health status and fault type of the fan blades. Acoustic signal detection is a non-contact, dynamic and non-destructive detection method suitable for long-term health monitoring of fan blades. Acoustic signal detection can monitor the health status of the fan blades in real time. The microphone does not need to be installed on the blades. It only needs to ensure that the original structure of the blades will not be damaged within the acoustic signal receiving range of the microphone. Moreover, installation and maintenance will not affect the normal operation of the wind turbine blades. Even if the microphone is damaged or fails, the wind turbine does not need to be shut down for replacement or repair, thus avoiding direct and indirect economic losses caused by shutdown due to maintenance of the monitoring system.

[0023] Furthermore, the vibration signal of the blade is collected from inside the blade; the sound signal of the blade is collected from outside the tower. This non-contact detection method will not cause damage to the structure, and installation and maintenance will not affect the normal operation of the wind turbine.

[0024] Furthermore, Fourier analysis is used to calculate the vibration signal to obtain the vibration frequency and vibration mode of the blade.

[0025] Furthermore, the Born-Jordan distribution is used to perform time-frequency analysis on the sound signal to obtain the abnormal whistle profile of the blade with high time and frequency resolution.

[0026] The present invention discloses a wind turbine blade early damage monitoring system based on vibration and sound. The system has a simple structure and has the characteristics of high detection efficiency, non-contact, low installation and operation and maintenance costs. It can reduce the probability of small damages gradually evolving into major accidents, reduce operating costs, and increase the number of available hours of the unit, thereby achieving the purpose of reducing production and operation costs.

[0027] Furthermore, the sound sensor uses a number of silicon microphones arranged in an array, which has high sensitivity, small size and good environmental adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the overall structure of the wind turbine blade early damage monitoring system based on vibration and sound of the present invention.

[0029] In the figure: 1- vibration signal acquisition module, 2- sound signal acquisition module, 3- noise reduction module, 4- processor. DETAILED DESCRIPTION

[0030] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments, the contents of which are intended to explain the present invention rather than to limit it:

[0031] like Figure 1 The wind turbine blade early damage monitoring system based on vibration and sound of the present invention comprises a vibration signal acquisition module 1, a sound signal acquisition module 2, a noise reduction module 3 and a processor 4; the vibration signal acquisition module 1 and the sound signal acquisition module 2 are respectively connected to the noise reduction module 3, and the noise reduction module 3 is connected to the processor 4. The processor 4 is connected to the remote data analysis platform through a communication module, and the communication module is preferably a wireless communication module.

[0032] The vibration signal acquisition module 1 includes a plurality of vibration sensors and a vibration signal collector. The plurality of vibration sensors are arranged and installed in the wind turbine blade cover plate. The vibration sensors are connected to the vibration signal collector, and the vibration signal collector is connected to the noise reduction module 3 .

[0033] The sound signal collection module 2 includes a sound sensor and a sound signal collector. The sound sensor is arranged outside the tower. The sound sensor is connected to the sound signal collector, and the sound signal collector is connected to the noise reduction module 3. The sound sensor is preferably a plurality of silicon microphones arranged in an array.

[0034] The method for monitoring early damage of wind turbine blades using the above-mentioned wind turbine blade early damage monitoring system based on vibration and sound comprises the following steps:

[0035] Step 1: respectively collect the vibration signal and sound signal of the blade through the vibration signal acquisition module 1 and the sound signal acquisition module 2; specifically including a vibration acceleration sensor, a sound acquisition microphone, a signal acquisition and processing device, a signal denoising module and an analysis and calculation module.

[0036] Step 2: Use the EEMD method to denoise the collected vibration and sound signals, and inject a white noise signal into the measured torsional vibration signal:

[0037] x m (t) = x(t) + n m (t)

[0038] The specific method is: let the original signal be x(t), and let r0(t)=x(t), i=1 under the initial conditions, and perform the following cyclic calculation:

[0039] (1) Let the function h0(t) = r i-1 (t), while j = 1;

[0040] (2) Find the intermediate function h j-1 The maximum and minimum points of (t) are interpolated using the cubic spline function to obtain h j-1 The upper envelope of (t) max (t) and the lower envelope e min (t), and the average of the upper and lower envelopes is: m j-1 (t) = [e max (t)+e min (t)] / 2, and finally get the next step intermediate function h j (t) = h j-1 (t)-m j-1 (t);

[0041] (3) Determine h based on the conditions that the above eigenmode function needs to satisfy j (t) Whether it is a modal function, if h j (t) is not an eigenmode function, then let j = j + 1 and continue with step (2) until h j (t) Satisfy the eigenmode function conditions;

[0042] (4) If h j (t) is an intrinsic mode function, then let imf i (t) = h j (t), r i (t) = r i-1 (t)-imf i (t), and judge r i (t) Is it monotonic or less than the set precision? If so, end the decomposition; otherwise, execute (1).

[0043] Step 3.1: Extract sound signal features. Perform Born-Jordan time-frequency analysis on the denoised sound signal x(t). To avoid confusion, rewrite x(t) as x(u). The Born-Jordan distribution is:

[0044]

[0045] Where f is the frequency, a is a parameter that is generally taken as 0.5, τ is the time-shift variable, * represents the conjugate function, and j is the imaginary unit.

[0046] At the time interval of the blade rotation cycle, a peak curve with similar morphology and continuous changes will appear. The peak curve with this morphological feature is defined as the characteristic morphological time-frequency curve of the blade fault. The time unit signal of the time-frequency curve is transformed into frequency-energy (decibel) and all peaks are searched; after filtering the searched peaks according to specific rules, the remaining peaks are combined into the original spectrum to form the outline of the abnormal whistle.

[0047] Step 3.2: Extract vibration signal features. Use Fourier analysis to calculate the blade's natural frequency and vibration mode from the vibration acceleration signal at each point. Compare and analyze the blade when it is newly installed, and calculate the change in each order of modal stiffness, so as to determine the degree of blade degradation and the cause of the failure:

[0048] Δf=f-f0

[0049]

[0050] where f i is a certain modal frequency calculated from the vibration signal using Fourier transform; f oi is the modal frequency of this order at the time of initial commissioning; M i is the modal mass corresponding to this order.

[0051] Step 4: Send the data analysis results obtained in steps 3.1 and 3.2 to the remote data analysis platform for diagnostic analysis. The remote data analysis platform has established a blade failure mode rule library and uses the IF...THE rule to obtain the blade failure mode.

[0052] The installation position of the present invention is further explained below in conjunction with a specific engineering example:

[0053] A hole is made in the cover plate of a wind turbine blade: a 25mm hole is made using an M25 hole opener; the vibration sensor is mounted on the surface of the blade and glued with aviation glue. After the vibration sensor is mounted, the cable is glued along the blade with AB glue, and the signal connector of the vibration sensor is led out from the hole in the cover plate.

[0054] The sound signal adopts a non-contact, dynamic and non-destructive detection method. The acoustic microphone array is used as a sound signal collection sensor, which is arranged at the bottom of the tower and connected to the wind turbine tower through a bracket.

[0055] To save space, the noise reduction module 3 and processor 4 are integrated in an industrial computer. An M9 hole is opened on the bracket between the pitch control for fixing the bottom plate. The industrial computer, socket, and communication module are all installed on the bottom plate. The industrial computer is fixed on the bottom plate at the bottom of the tower and directly installed on the tower. The installation deviation should meet the design or specification requirements. The cable must meet the power consumption and heat generation requirements of the industrial computer and communication module. This installation uses 1.5mm 2The cable is powered from the pitch control cabinet of the hub, and the cable length is about 2m. The connection head of the vibration sensor is led out from the blade cover and connected to the data acquisition interface of the industrial computer. To protect the signal cable, a 2m long spring is used to protect the signal cable to prevent wear and ensure long-term reliability and stability. The electrical cabinet in the cabin is installed on the ready-made guide rail on the side wall, fixed on the guide rail through the module slot, and the communication module is connected to the fiber optic converter through the network cable. The fiber optic converter uses a 110-meter-long single-mode optical fiber cable, and the optical fiber cable is tied together with the original power line of the wind turbine with a cable tie to connect to the ring network optical fiber network at the bottom of the tower. The server is deployed in the computer room of the central control room, and the installation and deployment can be completed by arranging it together with the existing server or switch, and it is privatized to ensure data security. By connecting to the local area network of the wind farm, the data can be transmitted to the computer monitor in the central control room. By entering the access address, account number, and password of the cloud platform monitoring system, the data of the acceleration sensors and acoustic emission sensors of the four wind turbines on site and the status of the wind turbine blades can be monitored in real time.

[0056] It should be noted that the above is only a part of the implementation of the present invention, and equivalent changes made to the system described in the present invention are all included in the protection scope of the present invention. Those skilled in the art of the present invention can replace the specific examples described in a similar manner, as long as they do not deviate from the structure of the present invention or exceed the scope defined by the claims, they all belong to the protection scope of the present invention.

Claims

1. A method for monitoring early damage of wind turbine blades based on vibration and sound, characterized in that: The following steps are involved: Step 1: Collect the vibration signal and sound signal of the blade; the vibration signal of the blade is collected from the inside of the blade; the sound signal of the blade is collected from the outside of the tower; Step 2: Perform noise reduction processing on the vibration signal and sound signal collected in step 1; The noise reduction process uses the EEMD method to inject white noise into the collected vibration signal and sound signal, and then perform empirical mode decomposition and signal reconstruction; In which, a white noise signal is injected into the measured torsional vibration signal: x m ( t )= x ( t ) + n m ( t ) The specific method is: suppose the original signal is x ( t ), and the initial condition is set r 0( t )= x ( t ), i =1, perform the following loop calculation: (1) Let the function h 0( t )= r i-1 ( t ),at the same time j =1; (2) Find the intermediate function h j-1 ( t )'s maximum and minimum points are interpolated using the cubic spline function to obtain h j-1 ( t )'s upper envelope e max ( t ) and the lower envelope e min ( t ), and the average of the upper and lower envelopes is: m j-1 ( t )=[ e max ( t )+ e min ( t )] / 2, and finally get the next intermediate function h j ( t )= h j-1 ( t )- m j-1 ( t ); (3) Judge based on the conditions that the eigenmode function needs to satisfy h j ( t ) is a modal function, if h j ( t ) is not an eigenmode function, then let j = j +1, continue with step (2) until h j ( t ) satisfies the eigenmode function condition; (4) If h j ( t ) is an eigenmode function, then let imf i ( t )= h j ( t ), r i ( t )= r i-1 ( t )- imf i ( t ), and judge r i ( t ) is monotonic or less than the set precision. If so, end the decomposition; otherwise, execute (1); Step 3: Extract features from the vibration signal after noise reduction in step 2 to obtain the vibration frequency and vibration mode of the blade; extract features from the sound signal after noise reduction to obtain the abnormal whistle profile of the blade; calculate the vibration signal using Fourier analysis to obtain the vibration frequency and vibration mode of the blade, and then compare and analyze the initial state of the blade when it is newly installed to calculate the change in stiffness of each order of modal stiffness; use Born-Jordan distribution to perform time-frequency analysis on the sound signal to obtain the abnormal whistle profile of the blade; Step 4: Monitor and analyze the status of the wind turbine blades based on the abnormal whistle profile, vibration frequency and vibration mode of the blades obtained in step 3; the data analysis results of the wind turbine blade status are sent to the remote data analysis platform for diagnostic analysis. The remote data analysis platform includes a blade failure mode rule library, and the blade failure mode is obtained based on the blade failure mode rule library.

2. A wind turbine blade early damage monitoring system based on vibration and sound, characterized in that: include A vibration signal acquisition module (1), a sound signal acquisition module (2), a noise reduction module (3) and a processor (4); the vibration signal acquisition module (1) and the sound signal acquisition module (2) are respectively connected to the noise reduction module (3), and the noise reduction module (3) is connected to the processor (4); the vibration signal acquisition module (1) and the sound signal acquisition module (2) are used to execute step 1 of the method according to claim 1, the noise reduction module (3) is used to execute step 2 of the method according to claim 1, and the processor (4) is used to execute steps 3 and 4 of the method according to claim 1.

3. The wind turbine blade early damage monitoring system based on vibration and sound according to claim 2 is characterized in that: The vibration signal collection module (1) comprises a plurality of vibration sensors and a vibration signal collector. The plurality of vibration sensors are arranged and installed in the wind turbine blade cover plate. The vibration sensors are connected to the vibration signal collector, and the vibration signal collector is connected to the noise reduction module (3).

4. The wind turbine blade early damage monitoring system based on vibration and sound according to claim 2 is characterized in that: The sound signal collection module (2) comprises a sound sensor and a sound signal collector, the sound sensor is arranged outside the tower, the sound sensor is connected to the sound signal collector, and the sound signal collector is connected to the noise reduction module (3).

5. The wind turbine blade early damage monitoring system based on vibration and sound according to claim 4 is characterized in that: The sound sensor is a number of silicon microphones arranged in an array.

Citation Information

Patent Citations

  • Wind generating set remote on-line multi-mode health state monitoring and fault diagnosis system

    CN103343728A

  • On-site acoustic diagnosis method and monitoring system for wind turbine blade failure

    CN104677623A

  • Wind turbine blade state monitoring method and system based on vibration detection and wireless signal transmission

    CN107061185A

  • Method for extracting and identifying abnormal whistle contour in wind sweeping sound signal of fan blade

    CN111161756A

  • Method and apparatus for monitoring a wind turbine

    US20200332775A1