WO2019119659A1 - 用于风力发电机组的涡激振动的监测方法和设备 - Google Patents

用于风力发电机组的涡激振动的监测方法和设备 Download PDF

Info

Publication number
WO2019119659A1
WO2019119659A1 PCT/CN2018/079570 CN2018079570W WO2019119659A1 WO 2019119659 A1 WO2019119659 A1 WO 2019119659A1 CN 2018079570 W CN2018079570 W CN 2018079570W WO 2019119659 A1 WO2019119659 A1 WO 2019119659A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
image
mark
edge
displacement
Prior art date
Application number
PCT/CN2018/079570
Other languages
English (en)
French (fr)
Inventor
杨博宇
Original Assignee
北京金风科创风电设备有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 北京金风科创风电设备有限公司 filed Critical 北京金风科创风电设备有限公司
Priority to AU2018282347A priority Critical patent/AU2018282347B2/en
Priority to EP18811125.6A priority patent/EP3530935B1/en
Publication of WO2019119659A1 publication Critical patent/WO2019119659A1/zh

Links

Images

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
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/30—Control parameters, e.g. input parameters
    • F05B2270/334—Vibration measurements
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2270/00—Control
    • F05B2270/80—Devices generating input signals, e.g. transducers, sensors, cameras or strain gauges
    • F05B2270/804—Optical devices
    • F05B2270/8041—Cameras

Definitions

  • the present disclosure relates to the field of wind power technology, and in particular to a method and apparatus for monitoring vortex induced vibration of a wind power generator set.
  • Wind power technology has now become a major contributor to the growing clean electricity market worldwide.
  • Wind turbines have many serious problems during the hoisting process, and vortex-induced vibration is one of the problems.
  • the occurrence of vortex-induced vibrations may cause an increase in the hoisting time of the unit and cause economic loss, and the occurrence of vortex-induced vibration may also cause a safety accident.
  • a spoiler device for the development of a vortex-induced vibration suppression device, a spoiler device is widely used and widely used.
  • the core functional components of the spoiler device are a spoiler block or a spoiler bar, and the spoiler block or the spoiler bar passes. Its own geometric features are attached to the surface of the tower, effectively controlling the separation position of the eddy currents, thereby achieving the effect of destroying the entire original regular flow field.
  • it is necessary to wind the spoiler on the upper 1/3 of the position during the hoisting process.
  • there is currently no better way to monitor the vortex-induced vibration nor can it The effect of monitoring the device for suppressing vortex-induced vibration is monitored.
  • the present disclosure provides a method and apparatus for monitoring vortex-induced vibration of a wind power generator, which realizes monitoring of vortex-induced vibration by collecting image data of a tower position of the wind power generator and processing the image data.
  • An aspect of the present disclosure provides a monitoring method for vortex induced vibration of a wind power generator, the monitoring method comprising the steps of: collecting image data of a tower of a wind power generator; and identifying the set in the tower from the image data An image of the mark at the predetermined position, and an image of the identified mark is tracked to obtain the displacement data of the mark; the vortex-induced vibration is monitored based on the displacement data of the mark.
  • Another aspect of the present disclosure also provides a monitoring system for vortex-induced vibration of a wind power generator, the monitoring system including the monitoring device and the collecting device for vortex-induced vibration of the wind power generator as described above, wherein
  • the acquisition device includes an image collector and a protection device.
  • Another aspect of the present disclosure provides a computer readable storage medium storing a computer program, the processor executing a monitoring method for vortex-induced vibration of a wind power generator as described above when the computer program is executed by a processor .
  • Another aspect of the present disclosure provides a computer device including a processor and a memory storing a computer program, the processor performing vortex-induced vibration for a wind power generator as described above when the computer program is executed by the processor Monitoring method.
  • the vortex-induced vibration of the wind turbine is monitored by means of image monitoring and template matching processing and edge detection processing on the image data, so that the wind turbine is modified by the wind turbine when hoisting
  • the hoisting time or the corresponding suppression device is added to avoid the occurrence of vortex-induced vibration, effectively avoiding economic losses and safety accidents, and improving the efficiency of wind turbine hoisting.
  • FIG. 1 shows a flow chart of a method for monitoring vortex-induced vibration of a wind power plant, in accordance with an embodiment of the present disclosure
  • FIG. 2 illustrates a schematic diagram of an installation manner of an image collector according to an exemplary embodiment of the present disclosure
  • FIG. 3 illustrates a flow diagram of data processing of acquired image data in accordance with an embodiment of the present disclosure
  • FIG. 4 illustrates a flowchart of calculating coordinate data of a marker 1 according to an exemplary embodiment of the present disclosure
  • FIG. 5 illustrates a block diagram of a monitoring apparatus for vortex-induced vibration of a wind power plant according to an embodiment of the present disclosure
  • FIG. 6 shows a block diagram of a data processing apparatus in accordance with an embodiment of the present disclosure
  • FIG. 7 is a block diagram showing a monitoring system for vortex induced vibration of a wind power generator set according to an embodiment of the present disclosure
  • FIG. 8 shows a schematic diagram of a protection device in accordance with an exemplary embodiment of the present disclosure.
  • FIG. 1 is a flowchart illustrating a monitoring method for vortex-induced vibration of a wind power generator set according to an embodiment of the present disclosure.
  • step S100 image data of a tower of a wind power generator is acquired.
  • an image collector such as a camera, is disposed at a distance from the wind power generator to capture the tower of the wind power generator, and the image data obtained by the photographing includes the photographing of two marks at an angle of 90 degrees.
  • Two sets of image data, the captured image data are acquired by two image collectors placed at an angle of 90 degrees.
  • one red rectangle mark may be brushed in advance in four different directions at the last tower position of the tower of the wind power generator, and any two of the four red rectangular marks may be selected.
  • the adjacent red rectangle marks 1 and 2 are taken, and the marks 1 and 2 are at an angle of 90 degrees.
  • the image collector shown in FIG. 2 is installed in such a manner that two cameras corresponding to the mark 1 and the mark 2 are respectively disposed at a position 200 meters away from the wind power generator, and the two cameras are placed at an angle of 90 degrees. And each of them takes an image of the tower corresponding to the marked position.
  • the above-described camera as an image collector is merely an illustrative example, and an image collector that can be employed in the present disclosure is not limited thereto.
  • the shape and color of the mark are merely exemplary, and other shapes and/or colors may be used.
  • step S200 an image of the marker set at a predetermined position of the tower is identified from the image data, and the image of the identified marker is tracked to acquire the displacement data of the marker.
  • template matching processing is performed on the collected image data to identify the sensitive area image of the marker, and then the edge detection processing is performed on the identified sensitive area image to obtain the edge contour data of the marker, and the marker is calculated based on the edge contour data of the marker.
  • Displacement data A process of recognizing an image of a marker set at a predetermined position of a tower from image data and tracking an image of the identified marker to acquire the displacement data of the marker according to an embodiment of the present disclosure will be described in detail below.
  • FIG. 3 is a flowchart illustrating data processing of acquired image data, according to an embodiment of the present disclosure.
  • step S201 the sensitive area image of the mark is identified by performing template matching processing on the acquired image data. Specifically, each frame of image data in a certain length of time is read, and then the template of the mark is respectively stacked on each frame of image data for template matching, and the matching degree of the template is judged by the correlation coefficient matching method, and then according to the template.
  • the degree of matching is used to identify the sensitive area image of the marker, wherein reading image data for a certain length of time can be read offline and/or online.
  • Template matching is to spatially align two or more images acquired by different sensors or the same sensor at the same time and under different imaging conditions, or to find the corresponding mode according to the known pattern. Processing method.
  • template matching is to search for a target in a large image. It is known that there are targets to be found in the image, and the target has the same size, direction and image as the template, and can be in the figure by a certain algorithm. Find the target and determine its coordinate position. According to the above example, each frame of image data within a certain length of time of the acquired marker 1 and marker 2 is read separately.
  • the image data of the captured marker 1 is taken as an example
  • the image data of the marker 1 is read in an online manner and the image data of each frame of the image data in the length of time t1-t2 is selected to obtain a series of images S h , wherein h is a positive integer greater than 0, and h represents the number of frames of the image data of the marker 1 in the time t1-t2.
  • h is a positive integer greater than 0
  • h represents the number of frames of the image data of the marker 1 in the time t1-t2.
  • a value of 100 for h indicates that the image data of the marker 1 is read for a length of time t1-t2 to obtain 100 image data.
  • the labeled templates are respectively stacked on the image S h for template matching.
  • the captured image S h is the searched image of the template matching process, and an intercepted image S h (W ⁇ H pixels) is arbitrarily selected as the searched image.
  • the template T is stacked on the S h for translation, and the template subgraph of the region of the searched image is Z ij , where i and j are the coordinates of the upper left corner of the region subgraph Z ij on the searched image, thereby knowing
  • the search range of the template matching is 1 ⁇ i ⁇ Wm, 1 ⁇ j ⁇ Hn, and the template matching process is completed by comparing the similarities of T and Z ij .
  • the correlation matching method is used to measure the matching degree of the template.
  • the numerator represents the product of two vectors
  • the denominator represents the modulo multiplication of two vectors.
  • the template matching processing is performed on the h images obtained by the above truncation, and h sensitive area images are obtained.
  • the template matching degree algorithm also includes a square difference matching method, a correlation matching method, a correlation coefficient matching method, a normalized matching method, and the like. It should be understood that the above correlation coefficient method is merely an exemplary example, and the template matching degree that can be adopted in the present disclosure. The algorithm is not limited to this.
  • edge contour data of the marker is acquired by performing edge detection processing on the identified sensitive area image. Specifically, the identified sensitive area image is filtered, the gradient of the filtered sensitive area image is obtained, and the non-maximum value suppression is performed according to the obtained gradient to perform edge enhancement on the sensitive area image, and then passes the threshold.
  • the algorithm performs edge detection on the edge-enhanced sensitive area image to obtain the marked edge contour data.
  • an sensitive area image c is arbitrarily selected for edge detection processing. Since the algorithm for edge detection of images is mainly based on the first and second derivatives of image intensity, and the derivative is usually very sensitive to noise, a filter is needed to improve the performance of the noise-related edge detector.
  • the Gaussian filtering method may be used to filter the sensitive area image c, that is, a discretized Gaussian function is used to generate a set of normalized Gaussian kernels, and then each of the image gray matrix is based on a Gaussian kernel function. One point is weighted and summed. Then, the edge enhancement of the filtered sensitive region image c is performed by calculating the gradient and the non-maximum value suppression, wherein the formula for calculating the magnitude and direction of the gradient is as follows:
  • G x and G y respectively represent the gradient sizes of different pixel positions on the image.
  • one of four possible angles generally 0 degree, 45 degree, 90 degree, 135 degrees
  • the maximum value of the pixel is found by non-maximum value suppression, that is, the pixel value of a certain point is determined in the field of 8 values. Whether it is the largest, the gray value corresponding to the non-maximum point is set to 0, thus eliminating the non-edge pixel points and achieving edge enhancement.
  • edge detection is performed on the edge-enhanced sensitive area image c by a threshold algorithm.
  • edge detection There are many methods for edge detection, such as Canny edge detection, Sobel edge detection, Roberts edge detection, Prewitt edge detection and Log edge detection.
  • Canny algorithm uses hysteresis threshold, which is double threshold method (high threshold and low threshold). .
  • edge detection is performed by the threshold algorithm, when the amplitude of a pixel position exceeds a high threshold, the pixel is reserved as an edge pixel. When the amplitude of a pixel position is less than a low threshold, the pixel is excluded. Point, when the amplitude of a pixel position is between a high threshold and a low threshold, then the pixel is only retained when connected to a pixel above a high threshold, and the remaining pixels are connected until the entire contour The edges are closed.
  • Step S202 performs edge detection processing on the h sensitive area images to obtain h edge contour data about the mark 1.
  • step S203 the displacement data of the marker is calculated based on the edge contour data of the marker.
  • the edge contour data of the marker is placed in the coordinate system, the pixel value of the pixel point of the edge contour is read, and the pixel value of the pixel of the next row is read when the pixel value is non-zero, thereby obtaining the edge coordinate of the marker.
  • the coordinate value of the coordinate of the mark is calculated, and the ordinate data of the coordinate average is recorded, and the difference data of the ordinate data of each frame of the recorded image is calculated to obtain the displacement data of the edge coordinate of the mark.
  • one edge contour data is arbitrarily selected from the h edge contour data obtained in step S202, the pixel value of each pixel point of the edge contour is read, and the pixel of the next row of pixel points is read when the pixel value is non-zero. Value, resulting in the edge coordinates of marker 1.
  • the coordinate data of the marker 1 is calculated based on the acquired h edge contour data according to an embodiment of the present disclosure, with reference to FIG.
  • FIG. 4 is a flowchart illustrating calculation of coordinate data of the marker 1 according to an exemplary embodiment of the present disclosure.
  • step S301 the pixel value of the pixel point of the Mth row and the Nth column is read. Assuming that one edge contour data f of the marker 1 is arbitrarily selected and placed in a Cartesian coordinate system, each pixel point x mn in f is read from the pixel point x 11 at the first row and the first column of the edge contour data f. The pixel value. In step S1, it is judged whether or not the pixel value is zero. When the pixel value is non-zero, step S302 is performed, and when the pixel value is zero, step S304 is performed.
  • step S302 the coordinate data of the pixel points of the Mth row and the Nth column are recorded, and if the pixel value of the pixel point x 11 of the edge contour data f is non-zero, the coordinate data (x 1 , y 1 ) of the pixel point is recorded. Then, step S303 is performed.
  • step S303 the pixel value of the pixel column of the Mth row and the Nth column is read. For example, the pixel value of the pixel point x 21 in the second row and the first column is read, and the pixel value is read and determined.
  • step S304 the pixel value of the pixel of the (N+1)th column of the Mth row is read. For example, if the pixel value of the pixel point x 11 is zero, the pixel value of the pixel point x 12 is read and judged.
  • the above-described operations of reading, judging the pixel value of the pixel and recording the coordinate data of the pixel are repeatedly performed, and the coordinate data of the pixel when the pixel value is non-zero is recorded, that is, the edge coordinate data of the marker 1 is obtained.
  • the above calculation is performed on the h edge contour data obtained in step S202, respectively, and the edge coordinate data of the h group mark 1 is obtained.
  • step S203 based on the obtained edge coordinate data of the marker 1, the left edge coordinate data or the right edge coordinate data of the marker 1 is extracted to calculate the coordinate average value, and the longitudinal value of the calculated coordinate average value is recorded.
  • Coordinate data Assume that the left edge coordinate data ⁇ (x 1 , y 1 ), (x 1 , y 2 ), (x 1 , y 3 ), ... (x 1 , y n ) ⁇ of the extracted marker 1 is obtained, and coordinates are obtained. The average is among them, Calculate the ordinate data of the average of h coordinates based on the obtained edge coordinate data of h group mark 1. Then, the difference calculation is performed on the h ordinate data to obtain the displacement data of the left edge coordinate of the marker 1.
  • the calculation of the above steps S201 to S203 is performed on the image data collected in step S100, respectively, and the displacement data of the left or right edge coordinates is obtained and the left or right edge coordinates obtained are obtained.
  • the displacement data is subjected to a displacement calculation to obtain the displacement data of the marker.
  • the image data of the collected marker 1 and the image data of the marker 2 are respectively calculated, and it is assumed that the displacement data of the left edge coordinates of the two markers obtained is with Displacement data with Perform displacement calculation to obtain the displacement data of the marker
  • step S300 the vortex induced vibration is monitored based on the displacement data of the mark.
  • the wind speed condition in which the vortex-induced vibration occurs is determined according to the displacement data of the mark and the wind speed data at the time of collecting the image data, and the increase suppression is judged based on the displacement data of the mark and the wind speed data at the time of collecting the image data. Whether the vortex-induced vibration of the wind turbine of the device occurs and the performance of the suppression device is evaluated.
  • the wind turbine does not use a vortex-induced vibration suppression device, and when the displacement data is marked
  • the wind turbine generates vortex-induced vibration
  • the wind speed condition in which the vortex-induced vibration occurs is determined according to the wind speed data when the image data is collected.
  • the hoisting operation time is changed or the corresponding vortex vibration suppression device is attached at the wind speed to perform hoisting.
  • the wind turbine is provided with a vortex-induced vibration suppression device if the displacement data of the mark If the condition is too large or the displacement direction changes rapidly, the performance of the suppression device added by the wind power generator is not good, and the suppression device needs to be inspected or replaced.
  • FIG. 5 is a block diagram illustrating a monitoring apparatus for vortex-induced vibration of a wind power generator set according to an embodiment of the present disclosure.
  • the monitoring device 400 for vortex-induced vibration of a wind turbine may include a data acquisition module 401, a data processing module 402, a vortex-induced vibration monitoring module 403, and an evaluation module 404.
  • the monitoring device 400 for vortex-induced vibration of a wind power generator may be implemented by various computing devices (eg, computers, servers, workstations, etc.).
  • the data acquisition module 401 is configured to acquire image data of a tower of a wind turbine
  • the data processing module 402 is configured to identify an image of the marker disposed at a predetermined position of the tower from the image data, and The image of the identified marker is tracked to obtain the displacement data of the marker
  • the vortex-induced vibration monitoring module 403 is configured to monitor the vortex-induced vibration based on the displacement data of the marker.
  • the evaluation module 404 is configured to determine wind speed conditions at which vortex-induced vibrations occur based on the displacement data of the markers and the wind speed data at which the image data is acquired. In addition, the evaluation module 404 further determines whether the vortex-induced vibration of the wind power generator to which the suppression device is added is based on the displacement data of the marker and the wind speed data when the image data is acquired, and performs performance evaluation on the suppression device.
  • a data processing module 402 in accordance with an embodiment of the present disclosure will be described in detail below with reference to FIG.
  • FIG. 6 shows a block diagram of a data processing module in accordance with an embodiment of the present disclosure.
  • the data processing module 402 includes a template matching unit 501, an edge detecting unit 502, and a displacement obtaining unit 503.
  • the template matching unit 501 identifies the sensitive area image of the mark by performing template matching processing on the acquired image data
  • the edge detecting unit 502 performs edge detection processing according to the sensitive area image recognized by the template matching unit 501, and acquires the edge contour of the mark.
  • the data, displacement obtaining unit 503 calculates the displacement data of the marker based on the edge contour data of the marker.
  • the template matching unit 501 reads the collected image data and intercepts each frame image data within a certain length of time, and then overlays the labeled templates on each frame image data for template matching, and judges by correlation coefficient matching method.
  • the degree of template matching which identifies the sensitive area image of the marker based on the degree of template matching.
  • the edge detecting unit 502 performs filtering processing on the identified sensitive area image, obtains a gradient of the filtered sensitive area image, and performs edge enhancement on the sensitive area image according to the obtained gradient for non-maximum value suppression, and then passes The threshold algorithm performs edge detection on the edge-enhanced sensitive area image to obtain the marked edge contour data.
  • common filtering methods mainly include a mean filtering method, a median filtering method, a bilateral filtering method, a Gaussian filtering method, a Wiener filtering method, etc., and the present disclosure uses a Gaussian filtering method to filter a sensitive area image.
  • the gradient value and direction of the sensitive area image obtained by the filtering process are calculated, and then the maximum value of the pixel is found by non-maximum value suppression according to the gradient direction, that is, whether the pixel value of a certain point is in the 8-value field.
  • edge enhancement of the sensitive area image is achieved.
  • the edge-enhanced sensitive area image is detected by the threshold algorithm, and the reserved pixel points are judged by the threshold value until the entire contour edge is closed, and the marked edge contour data is obtained.
  • the displacement obtaining unit 503 puts the edge contour data of the mark into the coordinate system, reads the pixel value of the pixel point of the edge contour, and reads the pixel value of the pixel of the next row when the pixel value is non-zero, thereby obtaining the edge of the mark Coordinates, then obtain the coordinate average value of the edge coordinates of the mark and record the ordinate data of the coordinate average value, and calculate the difference data of the ordinate data of each frame image data of the record to obtain the displacement data of the edge coordinate of the mark .
  • the acquired image data is two sets of image data obtained by capturing two marks at an angle of 90 degrees, and therefore, the edge coordinates of the marks of the two sets of image data are respectively obtained and extracted.
  • the left or right edge coordinate data of the mark is used to calculate the ordinate data of the coordinate average value, and then the displacement data of the coordinates of the two left or two right edges are obtained by calculating the difference value of the ordinate data.
  • the displacement data of the two left or two right edge coordinates obtained are combined and calculated to obtain the displacement data of the mark.
  • the evaluation module 404 is configured to determine the wind speed condition in which the vortex vibration occurs according to the displacement data of the mark and the wind speed data when the image data is acquired, and determine the increase suppression according to the displacement data of the mark and the wind speed data when the image data is collected. Whether the vortex-induced vibration of the wind turbine of the device occurs and the performance of the suppression device is evaluated. According to the embodiment of the present invention, for example, when the wind turbine does not add a suppressing device, the wind speed condition in which the vortex-induced vibration occurs is determined based on the displacement data of the mark and the wind speed data at the time of collecting the image data, and is implemented under the wind speed condition.
  • the lifting operation time is changed or the corresponding vortex-induced vibration suppression device is installed under the wind speed condition for lifting.
  • the wind turbine is provided with a suppressing device, it is judged whether or not the wind turbine generates vortex-induced vibration based on the displacement data of the mark and the wind speed data at the time of collecting the image data, and the performance of the suppressing device is evaluated. For example, when the displacement data of the mark is too large or the displacement direction changes rapidly, it is judged that the wind turbine generates vortex-induced vibration, and the performance of the suppression device added by the wind power generator is judged to be poor, and the suppression device is arranged. Check or replace.
  • FIG. 7 is a block diagram showing a monitoring system for vortex induced vibration of a wind power generator set according to an embodiment of the present disclosure.
  • the monitoring system 600 for vortex-induced vibration of a wind turbine includes a monitoring device 400 for vortex-induced vibration of the wind turbine and an acquisition device 601, wherein the acquisition device 601 is configured to collect wind power
  • the image data of the tower of the machine the monitoring device 400 is configured to monitor the vortex-induced vibration of the wind turbine by means of image monitoring and template matching processing and edge detection processing on the image data.
  • the acquisition device 601 includes an image capture device and a protection device for protecting the image capture device, wherein the protection device can protect the image collector with three sides, such as the protection device shown in FIG. It should be understood that the above-described trapezoidal baffle is merely an illustrative example, and the protective device that can be employed in the present disclosure is not limited thereto.
  • a method and apparatus for monitoring vortex-induced vibration of a wind power generator realizing a wind turbine generator by means of image monitoring and template matching processing and edge detection processing on image data
  • the monitoring of vortex-induced vibrations enables the wind turbines to avoid the occurrence of vortex-induced vibrations by changing the hoisting time of the wind turbines or adding corresponding suppression devices during hoisting, effectively avoiding economic losses and safety accidents, and improving wind power.
  • the monitoring method for vortex-induced vibration of a wind power generator may be implemented as computer readable code on a computer readable recording medium, or may be transmitted through a transmission medium.
  • the computer readable recording medium is any data storage device that can store data that can thereafter be read by a computer system.
  • the computer readable storage medium stores a computer program that, when executed by the processor, performs the monitoring method for vortex induced vibration of the wind turbine shown in FIG.
  • Examples of the computer readable recording medium include, but are not limited to, a read only memory (ROM), a random access memory (RAM), a compact disk (CD)-ROM, a digital versatile disk (DVD), a magnetic tape, a floppy disk, an optical data storage device.
  • Transmission media can include carriers that are transmitted over a network or various types of communication channels.
  • the computer readable recording medium can also be distributed over a computer system connected to the network such that the computer readable code is stored and executed in a distributed fashion.

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)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

本公开提供了一种用于风力发电机组的涡激振动的监测方法和设备，所述监测方法包括：采集风力发电机的塔筒的图像数据；从图像数据识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据；根据标记的位移数据对涡激振动进行监测。

Description

用于风力发电机组的涡激振动的监测方法和设备 技术领域
本公开涉及风电技术领域，具体地讲，涉及一种用于风力发电机组的涡激振动的监测方法和设备。
背景技术
风电技术现在已经成为全球日益增长的清洁电力市场的主要贡献。风力发电机在吊装过程中会出现很多严重的问题，涡激振动则是其中问题之一。涡激振动的发生会导致机组吊装时间的延长因而造成经济损失，并且涡激振动的发生也可能会造成安全事故发生。
现有技术中，对于涡激振动的抑制装置的开发，比较突出和广泛使用的是扰流装置，扰流装置的核心功能部件为扰流块或者扰流条，扰流块或者扰流条通过自身的几何特征依附在塔筒表面，有效的控制涡流的分离位置，从而达到将整个原始的规律流场破坏的效果。根据现有的塔架吊装工艺，在吊装过程中需要对其上1/3的位置缠绕扰流条，根据现有技术的限制，目前没有较好的方法可以对涡激振动进行监测，也无法对抑制涡激振动的装置进行效果监测。
发明内容
本公开提供了一种用于风力发电机组的涡激振动的监测方法和设备，通过采集风力发电机的塔筒位置的图像数据以及对图像数据进行处理，实现对涡激振动的监测。
本公开的一方面提供了一种用于风力发电机组的涡激振动的监测方法，所述监测方法包括以下步骤：采集风力发电机的塔筒的图像数据；从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据；根据标记的位移数据对涡激振动进行监测。
本公开的另一方面提供了一种用于风力发电机组的涡激振动的监测设备，所述监测设备包括：数据采集模块，被配置用于采集风力发电机的塔筒的图 像数据；数据处理模块，被配置用于从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据；涡激振动监测模块，被配置用于根据标记的位移数据对涡激振动进行监测。
本公开的另一方面还提供了一种用于风力发电机组的涡激振动的监测系统，所述监测系统包括如上所述的用于风力发电机组的涡激振动的监测设备和采集设备，其中，采集设备包括图像采集器和保护装置。
本公开的另一方面提供了一种计算机可读存储介质，存储有计算机程序，所述计算机程序被处理器运行时，处理器执行如上所述的用于风力发电机组的涡激振动的监测方法。
本公开的另一方面提供了一种包括处理器和存储计算机程序的存储器的计算机设备，当所述计算机程序被处理器运行时，处理器执行如上所述的用于风力发电机组的涡激振动的监测方法。
在本公开中，通过图像监测的方式和对图像数据进行模板匹配处理和边缘检测处理，实现了对风力发电机组的涡激振动的监测，使风力发电机组在进行吊装时，通过更改风力发电机组的吊装时间或者增设相应的抑制装置避免涡激振动的发生，有效的避免了经济损失和安全事故发生，提高了风力发电机组吊装的效率。
附图说明
通过以下结合附图进行的描述，本公开的示例性实施例的以上和其他方面、特点和优点将会更加清楚，在附图中：
图1示出根据本公开的实施例的用于风力发电机组的涡激振动的监测方法的流程图；
图2示出根据本公开的示例性的实施例的图像采集器的安装方式的示意图；
图3示出根据本公开的实施例的对采集的图像数据进行数据处理的流程图；
图4示出根据本公开的示例性的实施例的计算标记1的坐标数据的流程图；
图5示出根据本公开的实施例的用于风力发电机组的涡激振动的监测设备的框图；
图6示出根据本公开的实施例的数据处理装置的框图；
图7是示出根据本公开的实施例的用于风力发电机组的涡激振动的监测系统的框图；
图8示出根据本公开的示例性的实施例的保护装置的示意图。
在附图中，相同的标号将被理解为表示相同的元件、特征和结构。
具体实施方式
提供以下参照附图的描述以帮助全面理解由权利要求及其等同物限定的本公开的示例性实施例。以下参照附图的描述包括各种特定细节以帮助理解，但是所述特定细节将仅被视为示例性的。因此，本领域普通技术人员将意识到，在不脱离本公开的范围和精神的情况下，可对这里描述的实施例进行各种改变和修改。此外，为了清晰和简要，可省略公知功能和结构的描述。
图1是示出根据本公开的实施例的用于风力发电机组的涡激振动的监测方法的流程图。
如图1所示，首先，在步骤S100，采集风力发电机的塔筒的图像数据。具体地，在距离风力发电机一定距离处部署图像采集器，例如相机，对风力发电机的塔筒进行拍摄，拍摄所得的图像数据包括对呈90度夹角的两处标记进行拍摄所获取的两组图像数据，拍摄所得的图像数据通过呈90度夹角放置的两个图像采集器采集获取。根据本公开的实施例，可预先在风力发电机的塔筒的最后一节塔筒位置处的四个不同方向上各刷一个红色矩形标记，选择这四个红色矩形标记中的任意两个相邻的红色矩形标记1和2进行拍摄，且标记1和标记2呈90度夹角。例如图2中所示的图像采集器的安装方式，在距离风力发电机200米处的位置分别部署与标记1和标记2相对应的两台相机，则这两台相机呈90度夹角放置且各自拍摄对应标记位置的塔筒的图像。应理解，上述将相机作为图像采集器仅是示例性举例，本公开可采用的图像采集器不限于此。另外，标记的形状、颜色也仅是示例性举例，可采用其它形状和/或颜色的标记。
接下来，在步骤S200，从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据。具体地，对采集的图像数据进行模板匹配处理来识别标记的敏感区图像，然后，对识别的敏感区图像进行边缘检测处理以获取标记的边缘轮廓数据，并基于标记 的边缘轮廓数据来计算标记的位移数据。下面将参照图3来详细说明根据本公开实施例的从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据的过程。
图3是示出根据本公开的实施例的对采集的图像数据进行数据处理的流程图。
由图3可知，在步骤S201，通过对采集的图像数据进行模板匹配处理来识别标记的敏感区图像。具体地，读取一定时间长度内的每一帧图像数据，然后，将标记的模板分别叠放在每一帧图像数据上进行模板匹配，并通过相关系数匹配法判断模板匹配程度，再根据模板匹配程度来识别标记的敏感区图像，其中，读取一定时间长度内的图像数据可通过离线方式和/或在线方式进行读取。模板匹配就是把不同传感器或同一传感器在不同时间、不同成像条件下对同一景物进行采集获取的两幅或多幅图像在空间上对准，或者根据已知模式到另一幅图中寻找相应模式的处理方法。简而言之，模板匹配就是在一幅大图像中搜寻目标，已知该图中有要找的目标，且该目标同模板有相同的尺寸、方向和图像，通过一定的算法可以在图中找到目标，确定其坐标位置。根据上述举例，分别读取采集的标记1和标记2的一定时间长度内的每一帧图像数据。假设以采集的标记1的图像数据为例，通过在线方式读取标记1的图像数据并选取该图像数据在t1-t2时间长度内的每一帧图像数据得到一系列的图像S h，其中，h为大于0的正整数，h表示标记1的图像数据在t1-t2时间内的帧数。例如，h取值100，则表示在t1-t2时间长度内读取标记1的图像数据得到100个图像数据。然后，将标记的模板分别叠放在图像S h上进行模板匹配。假设标记的模板为模板T(m×n个像素)，截取的图像S h即是模板匹配处理的被搜索图，任意选取一张截取的图像S h(W×H个像素)作为被搜索图，将模板T叠放在S h上进行平移，模板覆盖被搜索图的区域子图为Z ij，其中，i、j为区域子图Z ij左上角在被搜索图上的坐标，由此可知，模板匹配的搜索范围是1≤i≤W-m，1≤j≤H-n，通过比较T和Z ij的相似性，完成模板匹配过程。最后，采用相关系数匹配法对模板匹配程度进行衡量。相关系数(r)是一种数学距离，可以用来衡量两个向量的相似程度，它起源于余弦定理cos(A)＝b 2+c 2-a 2/2bc，如果两个向量的夹角为0度(对应r＝1)，说明它们完全相似，如果夹角为90度(r＝0)，则它们完全不相似，如果夹角为180度(r＝-1)，则它们完全相反。把余弦定理写成向量的形式为：
cos(A)＝<b，c>/(|b|*|c|)
即：
Figure PCTCN2018079570-appb-000001
其中，分子表示两个向量的乘积，分母表示两个向量的模相乘。根据上式得出相关系数的求取公式如下：
Figure PCTCN2018079570-appb-000002
上式中，
Figure PCTCN2018079570-appb-000003
表示x j的平均值，
Figure PCTCN2018079570-appb-000004
表示y i的平均值。如果r＝1，则模板T和区域子图Z ij匹配结果完全相似，如果r＝0，则模板T和区域子图Z ij匹配结果完全不相似，由此判断出标记1的大致区域作为识别的敏感区图像。分别对上述截取得到的h个图像进行模板匹配处理，得出h个敏感区图像。模板匹配程度的算法还包括平方差匹配法、相关匹配法、相关系数匹配法和归一化匹配法等，应理解，上述相关系数法仅是示例性举例，本公开可采用的模板匹配程度的算法不限于此。
在步骤S202，通过对识别的敏感区图像进行边缘检测处理来获取标记的边缘轮廓数据。具体地，对识别的敏感区图像进行滤波处理，求取滤波处理后的敏感区图像的梯度，并根据求取的梯度进行非极大值抑制来对敏感区图像进行边缘增强，然后，通过阈值算法对边缘增强后的敏感区图像进行边缘检测以获取标记的边缘轮廓数据。根据上述举例，任意选取一个敏感区图像c对其进行边缘检测处理。由于对图像进行边缘检测的算法主要是基于图像强度的一阶和二阶导数，且导数通常对噪声很敏感，因此，需要采用滤波器来改善与噪声有关的边缘检测器的性能。常见的滤波方法主要有均值滤波法、中值滤波法、双边滤波法、高斯滤波法、维纳滤波法等。在本示例性实施例中可采用高斯滤波法对敏感区图像c进行滤波处理，即采用离散化的高斯函数产生一组归一化的高斯核，然后基于高斯核函数对图像灰度矩阵的每一点进行加权求和。然后，通过计算梯度和非极大值抑制对滤波处理后的敏感区图像c进行边缘增强，其中，计算梯度的辐值和方向的公式如下：
Figure PCTCN2018079570-appb-000005
Figure PCTCN2018079570-appb-000006
上式中，G x、G y分别表示图像上不同像素位置的梯度大小。根据梯度方向近似到四个可能角度之一(一般为0度、45度、90度、135度)，通过非极 大值抑制寻找像素点最大值，即判断某点像素值在8值领域内是否为最大，将非最大值点对应的灰度值设置为0，这样就排除了非边缘像素点，实现了边缘增强。最后，通过阈值算法对边缘增强后的敏感区图像c进行边缘检测。边缘检测的方法有很多种，例如Canny边缘检测、Sobel边缘检测、Roberts边缘检测、Prewitt边缘检测和Log边缘检测等，其中，Canny算法采用滞后阈值，也就是双阈值法(高阈值和低阈值)。采用阈值算法进行边缘检测时，当某一像素点位置的幅值超过高阈值时，则保留该像素点为边缘像素点，当某一像素点位置的幅值小于低阈值时，则排除该像素点，当某一像素点位置的幅值在高阈值和低阈值之间时，则该像素点仅仅在连接到一个高于高阈值的像素时被保留，把保留的像素点进行连接直到整个轮廓边缘闭合。步骤S202分别对h个敏感区图像进行边缘检测处理后得出关于标记1的h个边缘轮廓数据。
在步骤S203，基于标记的边缘轮廓数据来计算标记的位移数据。具体地，将标记的边缘轮廓数据放入坐标系中，读取边缘轮廓的像素点的像素值，在像素值非零时读取下一行像素点的像素值，由此得到标记的边缘坐标，然后再对标记的边缘坐标求取坐标平均值并记录坐标平均值的纵坐标数据，对记录的每一帧图像数据的纵坐标数据进行差值计算以得出标记的边缘坐标的位移数据。根据上述举例，在步骤S202得到的h个边缘轮廓数据中任意选取一个边缘轮廓数据，读取该边缘轮廓的各个像素点的像素值，并记录像素值非零时读取下一行像素点的像素值，得出标记1的边缘坐标。下面将参照图3来详细说明根据本公开实施例的基于获取的h个边缘轮廓数据来计算标记1的坐标数据。
图4是示出根据本公开的示例性的实施例的计算标记1的坐标数据的流程图。
由图4可知，在步骤S301中，读取第M行第N列像素点的像素值。假设任意选择标记1的一个边缘轮廓数据f并将其放入直角坐标系中，从边缘轮廓数据f的第1行第1列处的像素点x 11开始，读取f中各个像素点x mn的像素值。在步骤S1中，判断像素值是否为零。当像素值非零时，执行步骤S302，当像素值为零时，执行步骤S304。在步骤S302中，记录第M行第N列像素点的坐标数据，假设边缘轮廓数据f的像素点x 11的像素值非零，则记录下该像素点的坐标数据(x 1、y 1)，然后，执行步骤S303。
在步骤S303中，读取第M+1行第N列像素点的像素值。例如，读取第 2行第1列像素点x 21的像素值，对像素值进行读取、判断。
在步骤S304中，读取第M行第N+1列像素点的像素值。例如，像素点x 11的像素值为零，则读取像素点x 12的像素值并判断。
重复执行上述读取、判断像素点的像素值和记录像素点的坐标数据的操作，记录像素值非零时的像素点的坐标数据，即得出标记1的边缘坐标数据。分别对步骤S202中得到的h个边缘轮廓数据进行上述计算，则得到h组标记1的边缘坐标数据。
返回图3，在骤S203中，根据得到的标记1的边缘坐标数据，提取标记1左侧边缘坐标数据或者右侧边缘坐标数据来计算坐标平均值，并记录下计算得到的坐标平均值的纵坐标数据。假设提取的标记1的左侧边缘坐标数据{(x 1、y 1)、(x 1、y 2)、(x 1、y 3)...(x 1、y n)}，得出坐标平均值为
Figure PCTCN2018079570-appb-000007
其中，
Figure PCTCN2018079570-appb-000008
根据得到的h组标记1的边缘坐标数据计算得到h个坐标平均值的纵坐标数据
Figure PCTCN2018079570-appb-000009
然后，对h个纵坐标数据进行差值计算，得出标记1的左侧边缘坐标的位移数据。
根据本公开的实施例，分别对步骤S100中采集的图像数据进行上述步骤S201到步骤S203的计算，得出左侧或右侧边缘坐标的位移数据并将得出的左侧或右侧边缘坐标的位移数据进行合位移计算求取标记的位移数据。根据上述举例，分别对采集的标记1的图像数据和标记2的图像数据进行计算，假设求出的两个标记的左侧边缘坐标的位移数据为
Figure PCTCN2018079570-appb-000010
和
Figure PCTCN2018079570-appb-000011
则将位移数据
Figure PCTCN2018079570-appb-000012
和
Figure PCTCN2018079570-appb-000013
进行合位移计算求取标记的位移数据
Figure PCTCN2018079570-appb-000014
返回图1，在步骤S300，根据标记的位移数据对涡激振动进行监测。根据本公开的实施例，具体地，根据标记的位移数据和采集图像数据时的风速数据来确定发生涡激振动的风速条件，根据标记的位移数据和采集图像数据时的风速数据判断增设有抑制装置的风力发电机是否发生涡激振动，并对抑制装置进行性能评估。根据上述举例，假设该风力发电机没有采用涡激振动的抑制装置，则当标记的位移数据
Figure PCTCN2018079570-appb-000015
过大或者位移方向发生快速变化时，认为该风力发电机发生了涡激振动，再根据采集图像数据时的风速数据来确定发生涡激振动的风速条件。这样，在进行该风力发电机的吊装作业时，如果风速条件达到涡激振动的风速条件，则更改吊装作业的时间或者在该风速下加装相应的涡激振动的抑制装置进行吊装。假设该风力发电机增设有涡激振动的抑制装置，如果标记的位移数据
Figure PCTCN2018079570-appb-000016
过大或者位移方向发生快速变化，则该风 力发电机增设的抑制装置的性能不佳，需要对抑制装置进行排检或者更换抑制装置。
图5是示出根据本公开的实施例的用于风力发电机组的涡激振动的监测设备的框图。
如图5所示，用于风力发电机组的涡激振动的监测设备400可包括数据采集模块401、数据处理模块402、涡激振动监测模块403和评估模块404。根据本公开的实施例，用于风力发电机的对涡激振动的监测设备400可通过各种计算装置(例如，计算机、服务器、工作站等)来实现。具体的讲，数据采集模块401被配置用于采集风力发电机的塔筒的图像数据，数据处理模块402被配置用于从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据，涡激振动监测模块403被配置用于根据标记的位移数据对涡激振动进行监测。评估模块404被配置用于根据标记的位移数据和采集图像数据时的风速数据确定发生涡激振动的风速条件。此外，评估模块404还根据标记的位移数据和采集图像数据时的风速数据判断增设有抑制装置的风力发电机是否发生涡激振动，并对抑制装置进行性能评估。
下面将参照图6来详细说明根据本公开实施例的数据处理模块402。
图6示出根据本公开的实施例的数据处理模块的框图。
如图6所示，数据处理模块402包括模板匹配单元501、边缘检测单元502和位移求取单元503。其中，模板匹配单元501通过对获取的图像数据进行模板匹配处理来识别标记的敏感区图像，边缘检测单元502根据模板匹配单元501识别出的敏感区图像进行边缘检测处理，并获取标记的边缘轮廓数据，位移求取单元503基于标记的边缘轮廓数据来计算标记的位移数据。
模板匹配单元501读取采集的图像数据并截取一定时间长度内的每一帧图像数据，然后，将标记的模板分别叠放在每一帧图像数据上进行模板匹配，并通过相关系数匹配法判断模板匹配程度，根据模板匹配程度来识别标记的敏感区图像。根据本公开的实施例，模板匹配程度的算法由多种，例如平方差匹配法、相关匹配法、相关系数匹配法和归一化匹配法等，本公开采用的相关系数法中，当求取的相关系数r＝1时，标记的模板和所叠放的每一帧图像数据上重叠的区域几乎相似匹配，当r＝0时，标记的模板和所叠放的每一帧图像数据上重叠的区域完全不相似匹配，由此判断识别出标记的大致区域 作为敏感区图像。
边缘检测单元502对识别的敏感区图像进行滤波处理，求取滤波处理后的敏感区图像的梯度，并根据求取的梯度进行非极大值抑制来对敏感区图像进行边缘增强，然后，通过阈值算法对边缘增强后的敏感区图像进行边缘检测以获取标记的边缘轮廓数据。根据本公开的实施例，常见的滤波方法主要有均值滤波法、中值滤波法、双边滤波法、高斯滤波法、维纳滤波法等，本公开采用高斯滤波法对敏感区图像进行滤波处理。然后，滤波处理后对滤波处理得到的敏感区图像的梯度辐值和方向进行计算，再根据梯度方向通过非极大值抑制寻找像素点最大值，即判断某点像素值在8值领域内是否为最大，由此排除非边缘像素点，实现敏感区图像的边缘增强。最后，通过阈值算法检测边缘增强后的敏感区图像，并将通过阈值判断保留的像素点进行连接直到整个轮廓边缘闭合，得出标记的边缘轮廓数据。
位移求取单元503将标记的边缘轮廓数据放入坐标系中，读取边缘轮廓的像素点的像素值，在像素值非零时读取下一行像素点的像素值，由此得到标记的边缘坐标，然后对标记的边缘坐标求取坐标平均值并记录坐标平均值的纵坐标数据，对记录的每一帧图像数据的所述纵坐标数据进行差值计算得出标记的边缘坐标的位移数据。根据本公开的实施例，采集的图像数据是对呈90度夹角的两处标记进行拍摄获取的两组图像数据，因此，分别求取两组图像数据的标记的边缘坐标，并提取标记的左侧或者右侧边缘坐标数据来计算得出坐标平均值的纵坐标数据，然后，通过对纵坐标数据进行差值计算得出两个左侧或两个右侧边缘坐标的位移数据，并将得出的两个左侧或两个右侧边缘坐标的位移数据进行合位移计算求取标记的位移数据。
返回图5，评估模块404被配置用于根据标记的位移数据和采集图像数据时的风速数据确定发生涡激振动的风速条件，根据标记的位移数据和采集图像数据时的风速数据判断增设有抑制装置的风力发电机是否发生涡激振动，并对抑制装置进行性能评估。根据本发的实施例，例如，在风力发电机没有增设抑制装置时，根据标记的位移数据和采集图像数据时的风速数据来确定发生涡激振动的风速条件，并在该风速条件下对实施吊装作业的时间进行更改或者在该风速条件下加装相应的涡激振动的抑制装置进行吊装。或者假设在风力发电机增设有抑制装置时，根据标记的位移数据和采集图像数据时的风速数据判断风力发电机是否发生涡激振动，并以此对抑制装置进行性能评 估。例如，当标记的位移数据过大或者位移方向发生快速变化，则判断该风力发电机发生涡激振动，以及判断该风力发电机增设的抑制装置的性能不佳，并以此对抑制装置实施排检或者更换。
图7是示出根据本公开的实施例的用于风力发电机组的涡激振动的监测系统的框图。
如图7所示，用于风力发电机组的涡激振动的监测系统600包括用于风力发电机组的涡激振动的监测设备400和采集设备601，其中，采集设备601被配置用于采集风力发电机的塔筒的图像数据，监测设备400被配置用于通过图像监测的方式和对图像数据进行模板匹配处理和边缘检测处理，实现对风力发电机组的涡激振动的监测。采集设备601包括图像采集器和用于保护图像采集器的保护装置，其中，保护装置可采用呈梯形的挡板对图像采集器进行三面保护，例如图8所示的保护装置。应理解，上述呈梯形的挡板仅是示例性举例，本公开可采用的保护装置不限于此。
根据本公开的实施例的用于风力发电机组的涡激振动的监测方法和设备，该预测方法通过图像监测的方式和对图像数据进行模板匹配处理和边缘检测处理，实现了对风力发电机组的涡激振动的监测，使风力发电机组在进行吊装时，通过更改风力发电机组的吊装时间或者增设相应的抑制装置避免涡激振动的发生，有效的避免了经济损失和安全事故发生，提高了风力发电机组吊装的效率。
根据本公开的实施例的用于风力发电机组的涡激振动的监测方法可实现为计算机可读记录介质上的计算机可读代码，或者可通过传输介质被发送。计算机可读记录介质是可存储此后可由计算机系统读取的数据的任意数据存储装置。计算机可读存储介质存储有计算机程序，该计算机程序被处理器运行时，处理器执行图1所示的用于风力发电机组的涡激振动的监测方法。计算机可读记录介质的示例包括只读存储器(ROM)、随机存取存储器(RAM)、光盘(CD)-ROM、数字多功能盘(DVD)、磁带、软盘、光学数据存储装置，但不限于此。传输介质可包括通过网络或各种类型的通信通道发送的载波。计算机可读记录介质也可分布于连接网络的计算机系统，从而计算机可读代码以分布方式被存储和执行。
本公开的另一实施例提供了一种计算机设备，包括处理器和存储计算机程序的存储器，所述计算机程序被处理器运行时，处理器执行图1所示的用 于风力发电机组的涡激振动的监测方法。
尽管已经参照本公开的特定示例性实施例显示和描述了本公开，但是本领域技术人员将理解，在不脱离由权利要求及其等同物限定的本公开的精神和范围的情况下，可进行各种形式和细节上的各种改变。

Claims (21)

  1. 一种用于风力发电机组的涡激振动的监测方法，其特征在于，所述监测方法包括以下步骤：
    采集风力发电机的塔筒的图像数据；
    从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据；
    根据标记的位移数据对涡激振动进行监测。
  2. 如权利要求1所述的监测方法，其特征在于，所述从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据的步骤包括：
    通过对采集的图像数据进行模板匹配处理来识别标记的敏感区图像；
    通过对识别的敏感区图像进行边缘检测处理来获取标记的边缘轮廓数据；
    基于标记的边缘轮廓数据来计算标记的位移数据。
  3. 如权利要求2所述的监测方法，其特征在于，所述通过对采集的图像数据进行模板匹配处理来识别标记的敏感区图像的步骤包括：
    读取一定时间长度内的每一帧图像数据；
    将标记的模板分别叠放在所述每一帧图像数据上进行模板匹配；
    通过相关系数匹配法判断模板匹配程度，并根据模板匹配程度来识别标记的敏感区图像。
  4. 如权利要求2所述的监测方法，其特征在于，所述通过对识别的敏感区图像进行边缘检测处理来获取标记的边缘轮廓数据的步骤包括：
    对识别的敏感区图像进行滤波处理；
    求取滤波处理后的敏感区图像的梯度，并根据求取的梯度进行非极大值抑制来对敏感区图像进行边缘增强；
    通过阈值算法对边缘增强后的敏感区图像进行边缘检测以获取标记的边缘轮廓数据。
  5. 如权利要求2所述的监测方法，其特征在于，所述基于标记的边缘轮廓数据来计算标记的位移数据的步骤包括：
    将标记的边缘轮廓数据放入坐标系中，读取边缘轮廓的像素点的像素值，在像素值非零时读取下一行像素点的像素值，由此得到标记的边缘坐标；
    对标记的边缘坐标求取坐标平均值并记录坐标平均值的纵坐标数据，对记录的每一帧图像数据的纵坐标数据进行差值计算以得出标记的边缘坐标的位移数据。
  6. 如权利要求1-5中的任意一个所述的监测方法，其特征在于，所述图像数据包括对呈90度夹角的两处标记进行拍摄获取的两组图像数据，所述图像数据通过呈90度夹角放置的两个图像采集器采集获取。
  7. 如权利要求6所述的监测方法，其特征在于，所述基于标记的边缘轮廓数据来计算标记的位移数据的步骤还包括：将得出的两个左侧或两个右侧边缘坐标的位移数据进行合位移计算以求取标记的位移数据。
  8. 如权利要求1所述的监测方法，其特征在于，所述监测方法还包括：
    根据标记的位移数据和采集图像数据时的风速数据确定发生涡激振动的风速条件。
  9. 如权利要求1所述的监测方法，其特征在于，所述监测方法还包括：根据标记的位移数据和采集图像数据时的风速数据判断增设有抑制装置的风力发电机是否发生涡激振动，并对抑制装置进行性能评估。
  10. 一种用于风力发电机组的涡激振动的监测设备，其特征在于，所述监测设备包括：
    数据采集模块，被配置用于采集风力发电机的塔筒的图像数据；
    数据处理模块，被配置用于从图像数据中识别设置在塔筒预定位置处的标记的图像，并对识别的标记的图像进行追踪以获取标记的位移数据；
    涡激振动监测模块，被配置用于根据标记的位移数据对涡激振动进行监测。
  11. 如权利要求10所述的监测设备，其特征在于，所述数据处理模块包括：
    模板匹配单元，通过对采集的图像数据进行模板匹配处理来识别标记的敏感区图像；
    边缘检测单元，通过对识别的敏感区图像进行边缘检测处理来获取标记的边缘轮廓数据；
    位移求取单元，基于标记的边缘轮廓数据来计算标记的位移数据。
  12. 如权利要求11所述的监测设备，其特征在于，所述模板匹配单元被配置为：
    读取一定时间长度内的每一帧图像数据；
    将标记的模板分别叠放在所述每一帧图像数据上进行模板匹配；
    通过相关系数匹配法判断模板匹配程度，并根据模板匹配程度来识别标记的敏感区图像。
  13. 如权利要求11所述的监测设备，其特征在于，所述边缘检测单元被配置为：
    对识别的敏感区图像进行滤波处理；
    求取滤波处理后的敏感区图像的梯度，并根据求取的梯度进行非极大值抑制来对敏感区图像进行边缘增强；
    通过阈值算法对边缘增强后的敏感区图像进行边缘检测以获取标记的边缘轮廓数据。
  14. 如权利要求11所述的监测设备，其特征在于，所述位移求取单元被配置为：
    将标记的边缘轮廓数据放入坐标系中，读取边缘轮廓的像素点的像素值，在像素值非零时读取下一行像素点的像素值，由此得到标记的边缘坐标；
    对标记的边缘坐标求取坐标平均值并记录坐标平均值的纵坐标数据，对记录的每一帧图像数据的纵坐标数据进行差值计算以得出标记的边缘坐标的位移数据。
  15. 如权利要求10-14中的任意一个所述的监测设备，其特征在于，所述图像数据包括对呈90度夹角的两处标记进行拍摄获取的两组图像数据，所述图像数据通过呈90度夹角放置的两个图像采集器采集获取。
  16. 如权利要求15所述的监测设备，其特征在于，所述位移求取单元还被配置为：将得出的两个左侧或两个右侧边缘坐标的位移数据进行合位移计算以求取标记的位移数据。
  17. 如权利要求10所述的监测设备，其特征在于，所述监测设备还包括：
    评估模块，被配置用于根据标记的位移数据和采集图像数据时的风速数据确定发生涡激振动的风速条件以及根据标记的位移数据和采集图像数据时的风速数据判断增设有抑制装置的风力发电机是否发生涡激振动，并对抑制装置进行性能评估。
  18. 一种用于风力发电机组的涡激振动的监测系统，其特征在于，所述监测系统包括如权利要求10-17所述的用于风力发电机组的涡激振动的监测 设备和采集设备，其中，采集设备包括图像采集器和保护装置。
  19. 如权利要求18所述的监测系统，其特征在于，所述保护装置呈梯形对图像采集器进行三面保护。
  20. 一种计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器运行时，处理器执行权利要求1-9中任一项所述的监测方法。
  21. 一种计算机设备，包括处理器和存储计算机程序的存储器，其特征在于，所述计算机程序被处理器运行时，处理器执行如权利要求1-9中任一项所述的监测方法。
PCT/CN2018/079570 2017-12-18 2018-03-20 用于风力发电机组的涡激振动的监测方法和设备 WO2019119659A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
AU2018282347A AU2018282347B2 (en) 2017-12-18 2018-03-20 Method and apparatus for monitoring vortex-induced vibration of wind turbine
EP18811125.6A EP3530935B1 (en) 2017-12-18 2018-03-20 Method and equipment for monitoring vortex-induced vibration for wind turbine generator set

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711364848.2 2017-12-18
CN201711364848.2A CN109931229B (zh) 2017-12-18 2017-12-18 用于风力发电机组的涡激振动的监测方法和设备

Publications (1)

Publication Number Publication Date
WO2019119659A1 true WO2019119659A1 (zh) 2019-06-27

Family

ID=66982620

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/079570 WO2019119659A1 (zh) 2017-12-18 2018-03-20 用于风力发电机组的涡激振动的监测方法和设备

Country Status (4)

Country Link
EP (1) EP3530935B1 (zh)
CN (1) CN109931229B (zh)
AU (1) AU2018282347B2 (zh)
WO (1) WO2019119659A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113390376A (zh) * 2020-03-13 2021-09-14 新疆金风科技股份有限公司 确定风力发电机组的机舱位移的方法、装置和系统
CN113464380A (zh) * 2021-08-09 2021-10-01 观为监测技术无锡股份有限公司 一种塔筒安全性能确定方法、装置、设备及存储介质

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110296855B (zh) * 2019-07-05 2020-11-24 上海中认尚科新能源技术有限公司 一种评估风力发电机组高柔塔架抗涡激性能的检测方法
CN110486236B (zh) * 2019-08-08 2021-01-12 北京汉能华科技股份有限公司 一种风力发电机的故障检测方法和系统
CN110688906B (zh) * 2019-08-31 2023-07-18 深圳市广宁股份有限公司 风电设备的智能检测方法及相关产品
CN111735502B (zh) * 2020-07-16 2021-11-19 扬州大学 一种基于风场监测的涡振变流防护装置
CN111829747A (zh) * 2020-07-18 2020-10-27 扬州大学 一种基于固定翼无人机的沿海大桥涡激共振风场监控系统
CN112484649A (zh) * 2020-11-23 2021-03-12 北京汉风测控技术有限公司 一种基于图像识别的塔筒位移监测方法
CN113027701A (zh) * 2021-02-26 2021-06-25 中国电建集团华东勘测设计研究院有限公司 一种用于海上风机振动及冲刷试验的非接触式动态量测系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3002455A1 (de) * 2014-10-02 2016-04-06 RWE Innogy GmbH Verfahren und vorrichtung zum bestimmen von betriebsparametern einer windkraftanlage
CN105736252A (zh) * 2016-02-02 2016-07-06 华中科技大学 一种风电机组便携式点检记录仪
CN106677996A (zh) * 2016-12-29 2017-05-17 科诺伟业风能设备（北京）有限公司 一种风力发电机组塔筒振动异常检测方法
CN106895792A (zh) * 2017-04-27 2017-06-27 武汉大学 基于双目视觉测量系统的输电塔构件形变及内力监测方法

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020189376A1 (en) * 2000-07-24 2002-12-19 Markus Kalkum Procedure and device for the motion detection of microparticles
FR2852700B1 (fr) * 2003-03-19 2005-09-23 Centre Nat Rech Scient Procede et installation d'imagerie acousto-optique.
JP2006138862A (ja) * 2005-12-01 2006-06-01 Electron & Photon Laboratory Inc レーザ振動計
CN101196390A (zh) * 2007-12-19 2008-06-11 哈尔滨工业大学 基于视觉分辨的非接触式应变测量方法
CN101458119B (zh) * 2008-12-26 2010-08-11 常熟理工学院 图像分析实时检测轴瓦振动的方法
CN102134636A (zh) * 2010-01-26 2011-07-27 宝钢新日铁汽车板有限公司 监控退火炉内钢带跑偏的方法
CN102915545A (zh) * 2012-09-20 2013-02-06 华东师范大学 一种基于OpenCV的视频目标跟踪算法
CN103400379B (zh) * 2013-07-25 2016-08-31 国家电网公司 一种基于图像的覆冰检测和厚度估计方法
CN103837084B (zh) * 2014-02-18 2017-01-04 浙江华东工程安全技术有限公司 基于激光光斑成像技术的三向位移量测方法
KR101437920B1 (ko) * 2014-07-04 2014-11-04 세안기술 주식회사 카메라 이미지를 이용한 진동변위 측정 시 카메라 흔들림 보정방법
CN105352591A (zh) * 2015-12-04 2016-02-24 东华大学 一种纺纱锭子的振动特性测试方法
CN205352313U (zh) * 2016-01-08 2016-06-29 中铁第四勘察设计院集团有限公司 铁路构件位移监测系统
US10337495B2 (en) * 2016-05-19 2019-07-02 General Electric Company System and method for reducing vortex-induced tower vibrations of a wind turbine
CN106949936B (zh) * 2017-04-27 2019-10-11 武汉大学 利用双目视觉位移监测系统分析输电塔结构模态的方法
CN106989812B (zh) * 2017-05-03 2020-11-03 湖南科技大学 基于摄影测量技术的大型风机叶片模态测试方法
CN107314807B (zh) * 2017-07-07 2019-11-08 河南科技大学 一种钢丝绳横向空间振动测量的方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3002455A1 (de) * 2014-10-02 2016-04-06 RWE Innogy GmbH Verfahren und vorrichtung zum bestimmen von betriebsparametern einer windkraftanlage
CN105736252A (zh) * 2016-02-02 2016-07-06 华中科技大学 一种风电机组便携式点检记录仪
CN106677996A (zh) * 2016-12-29 2017-05-17 科诺伟业风能设备（北京）有限公司 一种风力发电机组塔筒振动异常检测方法
CN106895792A (zh) * 2017-04-27 2017-06-27 武汉大学 基于双目视觉测量系统的输电塔构件形变及内力监测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3530935A4 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113390376A (zh) * 2020-03-13 2021-09-14 新疆金风科技股份有限公司 确定风力发电机组的机舱位移的方法、装置和系统
CN113464380A (zh) * 2021-08-09 2021-10-01 观为监测技术无锡股份有限公司 一种塔筒安全性能确定方法、装置、设备及存储介质
CN113464380B (zh) * 2021-08-09 2022-09-27 观为监测技术无锡股份有限公司 一种塔筒安全性能确定方法、装置、设备及存储介质

Also Published As

Publication number Publication date
AU2018282347B2 (en) 2020-07-09
EP3530935A4 (en) 2019-12-18
CN109931229A (zh) 2019-06-25
EP3530935B1 (en) 2022-08-10
AU2018282347A1 (en) 2019-07-04
CN109931229B (zh) 2020-10-30
EP3530935A1 (en) 2019-08-28

Similar Documents

Publication Publication Date Title
WO2019119659A1 (zh) 用于风力发电机组的涡激振动的监测方法和设备
Lei et al. New crack detection method for bridge inspection using UAV incorporating image processing
CN109958583B (zh) 一种测量风力发电机组的塔架净空的方法和设备
TWI676963B (zh) 目標獲取方法及設備
Li et al. Road lane detection with gabor filters
CN107808161A (zh) 一种基于光视觉的水下目标识别方法
Phueakjeen et al. A study of the edge detection for road lane
CN106683040B (zh) 一种基于ncc算法的红外全景图像拼接方法
Sahoo et al. Design and simulation of various edge detection techniques using Matlab Simulink
Patel et al. 2D basic shape detection using region properties
Zhan et al. Effective waterline detection for unmanned surface vehicles in inland water
Shang et al. PolSAR land classification by using quaternion-valued neural networks
CN104715250A (zh) 十字激光检测方法和装置
CN112101134B (zh) 物体的检测方法及装置、电子设备和存储介质
CN107767366B (zh) 一种输电线路拟合方法及装置
Zheng et al. Research on edge detection algorithm in digital image processing
Hashmani et al. A survey on edge detection based recent marine horizon line detection methods and their applications
CN102262782A (zh) 一种利用cfa重采样插值定位拼接的数字图像取证方法
CN111062229B (zh) 用于风力发电机组外转子识别的方法及装置
Montesinos et al. A new perceptual edge detector in color images
JP6163732B2 (ja) 画像処理装置、プログラム、及び方法
CN113869163B (zh) 目标跟踪方法、装置、电子设备及存储介质
CN113469970A (zh) 一种涡激振动监测方法、系统、设备及存储介质
JP2011203853A (ja) 画像処理装置及びプログラム
JP6055307B2 (ja) 対応点探索装置およびカメラ姿勢推定装置ならびにこれらのプログラム

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2018282347

Country of ref document: AU

Date of ref document: 20180320

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18811125

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE