A fan operation fault intelligent troubleshooting system

Abstract

The application discloses a kind of fan operation fault intelligent troubleshooting system, it is related to the field of fault prediction and health management, solve the problem of insufficient accuracy of existing fan operation troubleshooting, including: data acquisition module: obtain operating state data and running observation data;Data processing module: operating state data is handled, the data range of operating state data is calculated, and data threshold is set;Running observation data is handled, the data variation of running observation data is calculated, and observation standard is obtained;Data analysis module: real-time monitoring is carried out to fan, the operating state of fan and observation standard are judged in combination with data threshold, analyze fan operation abnormal information, judge in combination with observation standard, analyze fan structure abnormal information;Fault early warning module: according to fan operation abnormality and fan structure abnormality, carry out fault early warning, and intelligently troubleshoot fan;The application can effectively improve the accuracy of fault troubleshooting, guarantee the operation safety of fan.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of fault prediction and health management, and more particularly to intelligent troubleshooting of fan operation faults, and specifically relates to a fan operation fault intelligent troubleshooting system. BACKGROUND

[0002] The existing fan operation fault has the following defects when troubleshooting:

[0003] 1. The detection direction of the existing fan operation fault intelligent troubleshooting system is limited, usually by analyzing historical data and analyzing existing fan operation data according to historical data. This analysis method has limited external environment perception, and single historical data can easily lead to data deviation, affecting troubleshooting results and reducing fault troubleshooting accuracy.

[0004] 2. The existing fan operation fault intelligent troubleshooting system has low fault warning interpretability. It is difficult to explain the diagnosis results through a black box model (such as a deep neural network), and the operation and maintenance personnel lack trust in the system decision, affecting fault handling efficiency and increasing manual inspection time.

[0005] 3. The existing fan operation fault intelligent troubleshooting system mainly focuses on troubleshooting key parts (such as bearings and gears), and the troubleshooting results are one-sided, making it difficult to effectively troubleshoot the entire fan. This reduces the accuracy and comprehensiveness of fault troubleshooting.

[0006] Therefore, we propose a fan operation fault intelligent troubleshooting system. SUMMARY

[0007] To address the deficiencies in the prior art, the application aims to provide a fan operation fault intelligent troubleshooting system that improves the accuracy of fan operation fault troubleshooting.

[0008] To achieve the above purpose, the application adopts the following technical solution: a fan operation fault intelligent troubleshooting system, and the specific working process of each module is as follows:

[0009] Data acquisition module: acquires fan operation data to obtain operation state data, and records fan operation images to obtain operation observation data;

[0010] Data processing module: processes the operation state data, calculates the data range of the operation state data, sets a data threshold according to the data range of the operation state data, and limits the operation state data; processes the operation observation data, calculates the data change of the operation observation data, and obtains the observation standard;

[0011] The data analysis module: real-time monitoring of the fan, obtaining monitoring data, combining data thresholds to judge the running state and observation standard of the fan, analyzing fan operation abnormal information, judging according to monitoring data combined with observation standards, analyzing fan structure abnormal information;

[0012] The fault early warning module: according to the fan operation abnormality and the fan structure abnormality, the fault early warning module carries out intelligent troubleshooting for the fan.

[0013] Further, the data acquisition module acquires data, specifically as follows:

[0014] The angular velocity of the blade is obtained, and the rotational speed of the blade is obtained according to the angular velocity of the blade. The rotational speed of the blade in different environments is obtained to obtain the blade rotational speed ypz(h). The rotational speed information of the blade is obtained by counting the rotational speed of the blade.

[0015] A plurality of measurement points are set on the blade, and the distance between each measurement point and the center line of the blade is obtained according to the number of measurement points to obtain the measurement distance clj(a). The vibration data of the blade is obtained by counting the measurement distance.

[0016] The number of metal particles in the oil of the fan is collected, the metal particle type bs is obtained, and each type of oil metal particle number yjk(b) is recorded according to the metal particle type bs. The wear information of the fan is obtained by counting the number of oil metal particles.

[0017] The running state data of the fan is composed of the rotational speed information of the blade, the vibration data of the blade and the wear information of the fan.

[0018] The surface image of the blade is collected to obtain the appearance image information of the blade. The thermal distribution of the blade is obtained, and the thermal image of the blade is constructed according to the thermal distribution of the blade. The running observation data of the fan is composed of the thermal image of the blade and the appearance image of the blade.

[0019] Further, the data is processed, specifically as follows:

[0020] According to the running state data of the fan, the rotational speed information of the blade, the vibration data of the blade and the wear information of the fan are obtained. The rotational speed information of the blade is processed to construct a rotational speed environment model, and the rotational speed of the blade is limited according to the rotational speed environment model. The vibration data of the blade is processed, the vibration threshold of the blade is set by calculating the vibration data under different rotating states. The wear information of the fan is processed, the number of oil metal particles is obtained from the wear information of the fan, and the wear threshold of the fan is obtained by calculating the number of oil metal particles.

[0021] According to the operation observation data of the fan, a thermal image of the blade and an appearance image of the blade are acquired, the thermal image of the blade is processed, thermal transfer data of the image is calculated, the thermal transfer standard is obtained by calculation according to the thermal transfer data of the image, the appearance image of the blade is processed, the gray value of the appearance image is acquired, and the gray value standard is obtained by calculation according to the gray value of the appearance image.

[0022] Further, the rotation speed information of the blade is processed, and the specific process is as follows:

[0023] According to the rotation speed information of the blade, the rotation speed ypz(h) of the blade is acquired, the rotation speed of the blades of cs different fans is acquired, and is recorded as ypz(h,c);

[0024] The rotation speeds ypz(h,c) of the blades of the fans under the same environmental influence are counted, the maximum rotation speed and the minimum rotation speed in the rotation speeds ypz(h,1) to ypz(h,cs) of the blades are extracted, the maximum rotation speed is recorded as zdz(h), and the minimum rotation speed is recorded as zxz(h);

[0025] The value of the environmental parameter h is acquired, the value of the environmental parameter h is recorded as h1 to hs, the maximum rotation speed and the minimum rotation speed of the fans under different environmental parameters are counted, and zdz(h1) to zdz(hs) and zxz(h1) to zxz(hs) are obtained;

[0026] According to the maximum rotation speed and the minimum rotation speed of the fans under different environmental parameters, a rotation speed environmental model hjm is constructed.

[0027] hjm=[mys(h)-zdz(h)]×[mys(h)-zxz(h)];

[0028] Wherein, mys(h) represents a preset rotation speed, which refers to an actual blade rotation speed under the environmental parameter h.

[0029] Further, the vibration data of the blade is processed, and the specific process is as follows:

[0030] According to the vibration data of the blade, the measurement distances at different angles when the blade rotates one circle are acquired, the measurement distances are discretized, d different angles are extracted, and the measurement distances are recorded as clj(a,d) according to the d different angles;

[0031] The distance between the measurement point and the center of the blade under the static condition is acquired, and a static distance jzj(a) is obtained; the measurement distance clj(a,d) is calculated according to the static distance jzj(a), the vibration angle of the blade is obtained, and the vibration threshold zyz(a) is obtained by mean calculation of the vibration angle of the blade.

[0032]

[0033] Further, the wear information of the fan is processed, and the specific process is as follows:

[0034] According to the wear information of the fan, the oil metal particle number yjk(b, c) of the cs different fans is obtained, the oil metal particles of each metal are counted according to the oil metal particle number yjk(b, c), the maximum value of the oil metal particle number is extracted, the maximum particle number zdk(b) is obtained, and the maximum particle number zdk(b) is taken as a single-class threshold value;

[0035] The oil metal particles of all kinds in each fan are counted to obtain the total number of oil metal particles of the fan, and the maximum value of the total number of oil metal particles is obtained to obtain a total number threshold value zsy.

[0036] The wear threshold value of the fan is composed of the single-class threshold value and the total number threshold value.

[0037] Further, the thermal image and the appearance image of the blade are processed, and the specific process is as follows:

[0038] According to the thermal image of the blade, the pixel number of the thermal image is obtained, and the pixel number of the image is is×js; the temperature of each pixel point is obtained to obtain the pixel temperature xwd(i, j); the pixel temperature xwd(i, j) is traversed and calculated to obtain the temperature transfer amount cd1(i, j) to cd4(i, j) in different directions;

[0039] According to the temperature transfer amount cd1(i, j) to cd4(i, j) in different directions, the transfer amount of all pixel temperatures on the thermal image is calculated to obtain a thermal transfer standard cdb.

[0040]

[0041] According to the appearance image of the blade, the gray value of the image is obtained, the data of the appearance image and the thermal image are aligned, the pixel number of the appearance image is is×js; according to the pixel number, the gray value of each pixel position is obtained to obtain the pixel gray value hdz(i, j); according to the pixel gray value, each row of the gray value of the appearance image of the blade is traversed to obtain the row gray value hhd(i);

[0042] According to the pixel gray value, each column of the gray value of the appearance image of the blade is traversed to obtain the column gray value lhd(j);

[0043] The row gray value hhd(i) and the column gray value lhd(j) are counted to obtain a gray value standard.

[0044] Further, the running state and the observation standard of the fan are judged, and the specific process is as follows:

[0045] The fan is monitored in real time to obtain fan monitoring data. According to the fan monitoring data, the real-time rotating speed of the fan is obtained. The real-time rotating speed is substituted into the rotating speed environment model for calculation. The rotating speed abnormality is analyzed. The real-time distance of the measuring point is obtained. According to the real-time distance of the measuring point, the vibration angle of the blade of the fan is calculated. In combination with the vibration threshold value, the vibration abnormality is analyzed. The real-time number of metal particles in the oil of the fan is obtained. In combination with the wear threshold value of the fan, the wear abnormality of the fan is analyzed. According to the rotating speed abnormality, the vibration abnormality and the wear abnormality of the fan, the operation abnormality information of the fan is formed.

[0046] According to the fan monitoring data, the real-time monitoring image of the fan is obtained. The real-time thermal image and the real-time appearance image of the fan are obtained through the real-time monitoring image of the fan. The real-time thermal image of the fan is processed in combination with the heat transfer standard to analyze the heat transfer abnormality of the fan. The real-time appearance image of the fan is processed in combination with the gray value standard to analyze the appearance abnormality of the fan. The structure abnormality information of the fan is formed by the heat transfer abnormality and the appearance abnormality of the fan.

[0047] Further, the operation abnormality information of the fan is obtained, specifically as follows:

[0048] The real-time rotating speed of the fan and the rotating speed environment model are obtained. The real-time rotating speed is substituted into the rotating speed environment model to obtain a rotating speed judgment value zpd. The rotating speed judgment value zpd is analyzed. If the rotating speed judgment value zpd is greater than 0, it indicates that the rotating speed of the fan is abnormal. If the rotating speed judgment value zpd is less than or equal to 0, it indicates that the rotating speed of the fan is normal.

[0049] The real-time distance of the measuring point is obtained. The real-time distance is denoted as ssj(a). The static distance jzj(a) is obtained. According to the real-time distance and the static distance, the real-time vibration angle is calculated. In combination with the vibration threshold value, a vibration judgment value zpd(a) is obtained.

[0050]

[0051] If the vibration judgment value zpd(a) is greater than 0, it indicates that the fan has a vibration abnormality, and the abnormal position is the position of the a-th measuring point. If the vibration judgment value zpd(a) is less than or equal to 0, it indicates that the fan vibration is normal.

[0052] The real-time number of metal particles in the oil of the fan is obtained. The real-time number of metal particles in the oil is denoted as syl(b). In combination with the wear threshold value of the fan, a total wear judgment value mpd and a single metal wear judgment value mpd(b) are obtained.

[0053]

[0054] Wherein mpd is the judging value of the total amount of all worn metal particles, mpd(b) is the wear judging of the bth metal particle; zsy is the total number threshold, zdk(b) is the single class threshold;

[0055] If mpd is greater than 0, it indicates that the fan is abnormally worn, if mpd(b) is greater than 0, it indicates that the bth metal particle in the fan is abnormally worn, and if mpd is less than or equal to 0 and mpd(b) is less than or equal to 0, it indicates that the fan is normally worn.

[0056] Further, the thermal transfer abnormality and the appearance abnormality of the fan are analyzed, and the specific process is as follows:

[0057] The real-time thermal image of the fan is acquired, the real-time temperature of different pixel positions is acquired according to the real-time thermal image of the fan, and is denoted as swd(i,j); the real-time temperature is traversed, and the thermal transfer standard cdb is combined for calculation to obtain a thermal transfer judging value rcp(i,j);

[0058] rcp(i,j) = |swd(i,j)-swd(i-1,j)| + |swd(i,j)-swd(i,j-1)|-2xcdb;

[0059] If the thermal transfer judging value rcp(i,j) is greater than 0, it indicates that the thermal transfer of the fan is abnormal, and the abnormal position is the position corresponding to the image pixel (i,j); if the thermal transfer judging value rcp(i,j) is less than 0, it indicates that the thermal transfer of the fan is normal.

[0060] The real-time appearance image of the fan is acquired, the gray value of each image pixel is acquired according to the real-time appearance image of the fan to obtain a real-time gray value shd(i,j); the real-time gray value is combined with the gray value standard for calculation to obtain a gray judging value;

[0061]

[0062] Wherein hdp(i) represents the gray value judging value of each row of pixels, ldp(j) represents the gray value judging value of each column of pixels; hhd(i) and lhd(j) are the gray value standards;

[0063] When hdp(i) is not equal to 0, it indicates that the i th row of pixels is abnormal, and when ldp(j) is not equal to 0, it indicates that the j th column of pixels is abnormal; the hdp(i) and ldp(j) are counted to obtain the appearance abnormality positioning.

[0064] As described above, due to the adoption of the above technical scheme, the present application has the following beneficial effects:

[0065] 1、The present application carries out fault troubleshooting to the fan through multi-source data, carries out data collection of multiple environments and multiple quantities to single data, guarantees the accuracy and anti-interference of data, guarantees the reliability of data analysis through the fixed mode of environmental variables, independently analyzes multiple data, troubleshoots the fan fault from multiple aspects, enhances the accuracy of fault troubleshooting, improves the safety and economy of the wind farm through timely troubleshooting and early warning of the fault.

[0066] 2、The present application carries out threshold setting and division standard to data, carries out calculation through data threshold and data standard combined with real-time data of the fan, accurately judges the abnormal condition of the real-time data of the fan, accurately locates the abnormality, and improves the efficiency of abnormal processing.

[0067] 3、The present application carries out comprehensive analysis of the fan through the running state data and running observation data of the fan, carries out comprehensive analysis of the parts, joints and surface of the fan, independently analyzes different parts, enhances the troubleshooting accuracy of single direction, carries out comprehensive analysis of the whole, and guarantees the comprehensive troubleshooting of the fan. BRIEF DESCRIPTION OF DRAWINGS

[0068] In order to facilitate the understanding of those skilled in the art, the present application will be further described below in combination with the drawings.

[0069] Figure 1 It is the overall system block diagram of the present application;

[0070] Figure 2 It is the schematic diagram of the measurement point selection in the present application;

[0071] Figure 3 It is the data processing schematic diagram in the present application;

[0072] Figure 4 It is the vibration calculation schematic diagram in the present application; DETAILED DESCRIPTION

[0073] The technical solutions of the present application will be described below in combination with the embodiments, obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the present application.

[0074] Embodiment one

[0075] Please refer to Figure 1The application belongs to the field of fault prediction and health management, and provides a technical solution: a fan operation fault intelligent troubleshooting system, comprising a data acquisition module, a data processing module, a data analysis module, a fault early warning module and a server, wherein the data acquisition module, the data processing module, the data analysis module and the fault early warning module are connected with the server, and the server controls the data acquisition module, the data processing module, the data analysis module and the fault early warning module respectively.

[0076] The data acquisition module: acquires the operation data of the fan to obtain the operation state data, and records the operation image of the fan to obtain the operation observation data.

[0077] It should be noted that the fan in the application mainly refers to a wind turbine, wherein the operation state data mainly includes the rotating speed of the blade, the vibration condition of the blade rotation and the wear condition during operation; and the operation observation data mainly includes the physical state of the fan, including the surface scratch and coating separation of the fan, and the conduction temperature of the fan blade.

[0078] The specific working process of the data acquisition module is as follows:

[0079] The angular velocity of the blade is acquired through the angular velocity sensor, and the rotating speed of the blade is obtained according to the angular velocity of the blade; the rotating speed of the blade under different environments is acquired to obtain the rotating speed ypz(h) of the blade; and the rotating speed information of the blade is obtained by counting the rotating speed of the blade.

[0080] It should be noted that the different environments mainly refer to different wind influences; and ypz(h) represents the rotating speed of the blade under the influence of the environment parameter h.

[0081] Please refer to Figure 2 A plurality of measurement points are set on the blade, the distance between each measurement point and the center line of the blade is acquired according to the number of measurement points to obtain the measurement distance clj(a); and the vibration data of the blade is obtained by counting the measurement distance.

[0082] It should be noted that the center line of the blade refers to a straight line passing through the center point of the blade and perpendicular to the surface of the blade.

[0083] The number of oil metal particles of the fan is acquired through the oil sensor, the metal particle category bs is acquired, each oil metal particle number yjk(b) is recorded according to the metal particle category bs; and the wear information of the fan is obtained by counting the oil metal particle number.

[0084] The operation state data of the fan is composed of the rotating speed information of the blade, the vibration data of the blade and the wear information of the fan.

[0085] The appearance image information of the blade is obtained by collecting the blade surface image through an optical camera, the thermal distribution of the blade is obtained through infrared thermal imaging, and the thermal image of the blade is constructed according to the thermal distribution of the blade; the operation observation data of the fan is formed according to the thermal image of the blade and the appearance image of the blade;

[0086] The data processing module processes the operation state data, calculates the data range of the operation state data, sets a data threshold according to the data range of the operation state data, and limits the operation state data; the operation observation data is processed, the data change of the operation observation data is calculated, and the observation standard is obtained;

[0087] The specific working process of the data processing module is as follows:

[0088] According to the operation state data of the fan, the rotation speed information of the blade, the vibration data of the blade and the wear information of the fan are obtained, the rotation speed information of the blade is processed, a rotation speed environment model is constructed, and the rotation speed of the blade is limited according to the rotation speed environment model; the vibration data of the blade is processed, the vibration threshold of the blade is set by calculating the vibration data under different rotation states; the wear information of the fan is processed, the number of metal particles in oil is obtained from the wear information of the fan, the number of metal particles in oil is calculated, and the wear threshold of the fan is obtained;

[0089] According to the operation observation data of the fan, the thermal image of the blade and the appearance image of the blade are obtained, the thermal image of the blade is processed, the thermal transfer data of the image is calculated, the thermal transfer standard is obtained by calculating the thermal transfer data of the image, the appearance image of the blade is processed, the gray value of the appearance image is obtained, and the gray value standard is obtained by calculating the gray value of the appearance image;

[0090] Please refer to Figure 3 According to the rotation speed information of the blade, the blade rotation speed ypz(h) is obtained, the blade rotation speeds of cs different fans are obtained, denoted as ypz(h,c), ypz(h,c) represents the blade rotation speed of the cth fan under the influence of the environment parameter h;

[0091] It should be noted that: the cs different fans refer to the fan in normal operation, and the blade rotation speed of the fan under normal operation is judged by obtaining the blade rotation speed of multiple fans.

[0092] The blade rotation speeds ypz(h,c) of the fans under the same environmental influence are counted, and the maximum rotation speed and the minimum rotation speed in the blade rotation speeds ypz(h,1) to ypz(h,cs) are extracted; the maximum rotation speed is denoted as zdz(h), and the minimum rotation speed is denoted as zxz(h);

[0093] Obtain the value of the environmental parameter h, and record the value of the environmental parameter h as h1 to hs; the maximum and minimum speeds of the fan under different environmental parameters are counted to obtain zdz(h1) to zdz(hs) and zxz(h1) to zxz(hs);

[0094] According to the maximum and minimum speeds of the fan under different environmental parameters, a speed environmental model hjm is constructed;

[0095] hjm = [mys(h)-zdz(h)]x[mys(h)-zxz(h)];

[0096] Wherein, mys(h) represents the preset speed, which refers to the actual blade speed under the environmental parameter h;

[0097] It should be noted that the maximum and minimum speeds of the fan under different environmental parameters are used to limit the range of the actual speed of the fan, ensuring the normal speed of the fan blades; the interval speed and the interval speed outside the interval are divided by the environmental model, which is convenient for extracting abnormal speed.

[0098] Please refer to Figure 4 According to the vibration data of the blade, the measurement distance at different angles when the blade rotates one circle is obtained, the measurement distance is discretized, and ds different angles are extracted; according to the ds different angles, the measurement distance is recorded as clj(a, d); clj(a, d) represents the measurement distance of the a-th measurement point at the d-th angle;

[0099] Obtain the distance between the measurement point and the center of the blade under static condition to obtain the static distance jzj(a); calculate the measurement distance clj(a, d) according to the static distance jzj(a) to obtain the vibration angle of the blade, and calculate the mean value of the vibration angle of the blade to obtain the vibration threshold zyz(a);

[0100]

[0101] It should be noted that the deflection angle of the blade when rotating is calculated to reflect the vibration of the blade, and the vibration information of different blade positions is independently processed by multi-point measurement, which is convenient for accurately extracting the abnormal position of the blade vibration.

[0102] According to the wear information of the fan, the oil metal particle number yjk(b, c) of cs different fans is obtained, the oil metal particle number of each metal oil is counted according to the oil metal particle number yjk(b, c), the maximum value of the oil metal particle number is extracted, and the maximum particle number zdk(b) is obtained as a single threshold;

[0103] It should be noted that: the maximum particle number zdk(b) refers to the maximum value corresponding to the b-th kind of particle in the oil metal particle number yjk(b,c) of the cs different fans, which reflects the maximum value of the wear of the metal particle under normal circumstances; the single-class threshold value represents the value threshold of the metal particles of a certain kind.

[0104] The total number of oil metal particles of the fan is obtained by counting all kinds of oil metal particles in each fan, and the maximum value of the total number of oil metal particles is obtained to obtain a total number threshold zsy.

[0105] The wear threshold of the fan is formed by the single-class threshold value and the total number threshold value.

[0106] According to the thermal image of the blade, the number of pixels of the thermal image is obtained, and the number of pixels of the image is denoted as is x js; the temperature of each pixel point is obtained to obtain the pixel temperature xwd(i,j); xwd(i,j) represents the temperature of the pixel point in the i-th row and the j-th column; the pixel temperature xwd(i,j) is traversed, and the difference value calculation is performed on the pixel temperature xwd(i+1,j), the pixel temperature xwd(i-1,j), the pixel temperature xwd(i,j+1) and the pixel temperature xwd(i,j-1) respectively, to obtain the temperature transfer amount cd1(i,j) to cd4(i,j) in different directions.

[0107] cd1(i,j) = |xwd(i,j)-xwd(i+1,j)|;

[0108] cd2(i,j) = |xwd(i,j)-xwd(i-1,j)|;

[0109] cd3(i,j) = |xwd(i,j)-xwd(i,j+1)|;

[0110] cd4(i,j) = |xwd(i,j)-xwd(i,j-1)|;

[0111] The transfer amount of all pixel temperatures on the thermal image is calculated according to the temperature transfer amount cd1(i,j) to cd4(i,j) in different directions to obtain a thermal transfer standard cdb.

[0112]

[0113] It should be noted that: in the fan blade, heat is quickly conducted through continuous material layers, and the surface temperature is uniformly distributed. By calculating the uniformly distributed temperature, the temperature error is accurately calculated, the error standard of temperature transfer is set, and the blade anomaly is accurately extracted through the error standard.

[0114] According to the appearance image of the blade, the gray value of the image is obtained, the appearance image is data-aligned with the thermal image, so that the pixel number of the appearance image is is x js; according to the pixel number, the gray value of each pixel position is obtained, and the pixel gray value hdz(i, j) is obtained; according to the pixel gray value, each row gray value of the appearance image of the blade is traversed, and the row gray value hhd(i) is obtained;

[0115]

[0116] According to the pixel gray value, each column gray value of the appearance image of the blade is traversed, and the column gray value lhd(j) is obtained;

[0117]

[0118] The row gray value hhd(i) and the column gray value lhd(j) are counted, and the gray value standard is obtained.

[0119] It should be noted that by counting the rows and columns of the pixel gray value, the two-dimensional positioning of the abnormal gray value is realized, and the positioning processing steps are simplified.

[0120] The data analysis module: real-time monitoring of the fan, obtaining monitoring data, combining data threshold to judge the running state of the fan and observation standard, analyzing fan running abnormal information, combining observation standard to judge according to monitoring data, analyzing fan structure abnormal information;

[0121] The specific working process of the data analysis module is as follows:

[0122] The fan is monitored in real time, the monitoring data of the fan is obtained, the real-time speed of the fan is obtained according to the monitoring data of the fan, the real-time speed is substituted into the speed environment model for calculation, the speed abnormality is analyzed, the real-time distance of the measuring point is obtained, the vibration angle of the blade of the fan is calculated according to the real-time distance of the measuring point, the vibration threshold is combined, the vibration abnormality is analyzed, the real-time oil metal particle number of the fan is obtained, the wear threshold of the fan is combined, and the wear abnormality of the fan is analyzed; according to the speed abnormality, vibration abnormality and wear abnormality of the fan, the running abnormal information of the fan is formed;

[0123] According to the monitoring data of the fan, the real-time monitoring image of the fan is obtained, the real-time thermal image and the real-time appearance image of the fan are obtained through the real-time monitoring image of the fan, the real-time thermal image of the fan is combined with the heat transfer standard for processing, and the heat transfer abnormality of the fan is analyzed; the real-time appearance image of the fan is combined with the gray value standard for processing, and the appearance abnormality of the fan is analyzed; the heat transfer abnormality and the appearance abnormality of the fan are combined to form the structure abnormal information of the fan;

[0124] The real-time rotating speed of the fan is acquired, and a rotating speed environmental model is acquired, the real-time rotating speed is substituted into the rotating speed environmental model to obtain a rotating speed judgment value zpd, and the rotating speed judgment value zpd is analyzed; if the rotating speed judgment value zpd is greater than 0, it indicates that the rotating speed of the fan is abnormal, and if the rotating speed judgment value zpd is less than or equal to 0, it indicates that the rotating speed of the fan is normal;

[0125] The real-time distance of the measuring point is acquired, and the real-time distance is recorded as ssj(a), the static distance jzj(a) is acquired, the real-time vibration angle is calculated according to the real-time distance and the static distance, the vibration threshold is combined for calculation, and a vibration judgment value zpd(a) is obtained;

[0126]

[0127] If the vibration judgment value zpd(a) is greater than 0, it indicates that the fan has vibration abnormality, and the abnormal position is the position of the a-th measuring point; if the vibration judgment value zpd(a) is less than or equal to 0, it indicates that the fan vibration is normal.

[0128] The real-time oil metal particle number of the fan is acquired, the real-time oil metal particle number is recorded as syl(b), the wear threshold of the fan is combined for calculation, and a total wear judgment value mpd and a single metal wear judgment value mpd(b) are obtained;

[0129]

[0130] Wherein mpd is the judgment value of the total amount of all wear metal particles, mpd(b) is the wear judgment of the b-th metal particle; zsy is the total number threshold, and zdk(b) is the single type threshold;

[0131] If mpd is greater than 0, it indicates that the fan has wear abnormality, if mpd(b) is greater than 0, it indicates that the b-th metal particle in the fan has wear abnormality, and if mpd is less than or equal to 0 and mpd(b) is less than or equal to 0, it indicates that the fan has normal wear.

[0132] The real-time thermal image of the fan is acquired, the real-time temperature of different pixel positions is acquired according to the real-time thermal image of the fan, and is recorded as swd(i,j); the real-time temperature is traversed, the heat transfer standard cdb is combined for calculation, and a heat transfer judgment value rcp(i,j) is obtained;

[0133] rcp(i,j) = |swd(i,j)-swd(i-1,j)| + |swd(i,j)-swd(i,j-1)|-2×cdb;

[0134] If the heat transfer judgment value rcp(i, j) is greater than 0, it indicates that the fan heat transfer is abnormal, and the abnormal position is the position corresponding to the image pixel (i, j); if the heat transfer judgment value rcp(i, j) is less than 0, it indicates that the fan heat transfer is normal.

[0135] An image of the real-time appearance of the fan is obtained, the gray value of each image pixel is obtained according to the real-time appearance image of the fan, and a real-time gray value shd(i, j) is obtained; the gray value judgment value is obtained by combining the gray value standard with the real-time gray value;

[0136]

[0137] Where hdp(i) represents the gray value judgment value of each row of pixels, and ldp(j) represents the gray value judgment value of each column of pixels; hhd(i) and lhd(j) are gray value standards;

[0138] When hdp(i) is not equal to 0, it indicates that the i-th row of pixels is abnormal, and when ldp(j) is not equal to 0, it indicates that the j-th column of pixels is abnormal; the appearance abnormality is located according to the statistics of hdp(i) and ldp(j).

[0139] It should be noted that: by separately analyzing the speed information of the blade, the vibration data of the blade, the wear information of the fan, the thermal image of the blade and the appearance image of the blade, the accuracy of the fan judgment is enhanced, and multiple directions are analyzed to ensure the comprehensiveness of the analysis results.

[0140] The fault warning module: according to the fan operation abnormality and the fan structure abnormality, the fault warning is carried out, and the intelligent fault troubleshooting of the fan is carried out;

[0141] The specific working process of the fault warning module is as follows:

[0142] Obtain the running abnormal information of the fan, obtain the speed abnormality, vibration abnormality and wear abnormality of the fan according to the running abnormal information of the fan, obtain the analysis information of the speed abnormality, vibration abnormality and wear abnormality from the speed abnormality, vibration abnormality and wear abnormality of the fan, upload the abnormality and analysis information, and maintain the maintenance personnel;

[0143] Obtain the structure abnormal information of the fan, obtain the heat transfer abnormality and appearance abnormality of the fan according to the structure abnormal information of the fan, transmit the heat transfer abnormality to the maintenance personnel for analysis, make abnormality judgment, maintain the fan, and make abnormality positioning of the appearance abnormality, and warn the maintenance personnel.

[0144] The preferred embodiments of the application disclosed above are only to facilitate the elucidation of the application. The preferred embodiments do not describe all the details of the application and limit the application to the specific embodiments. Obviously, many modifications and variations can be made in light of the teachings above. The description is chosen and described in order to provide the best illustration of the application and its practical application to those skilled in the art and to enable those skilled in the art to best utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A wind turbine operation fault diagnosis system, characterized in that, include: Data acquisition module: Collects operating data of the wind turbine to obtain operating status data, records operating images of the wind turbine to obtain operating observation data; Data processing module: processes operational status data, calculates the data range of operational status data, sets data thresholds based on the data range of operational status data, and limits the operational status data; processes operational observation data, calculates the data changes of operational observation data, and obtains observation standards; The data is processed as follows: Based on the operating status data of the wind turbine, the blade rotation speed, blade vibration data, and wind turbine wear information are obtained. The blade rotation speed information is processed to construct a rotation speed environment model, and the blade rotation speed is limited according to the rotation speed environment model. The blade vibration data is processed, and the vibration threshold of the blade is established by calculating the vibration data under different rotation states. The wind turbine wear information is processed, and the number of oil metal particles is obtained from the wind turbine wear information. The wear threshold of the wind turbine is obtained by calculating the number of oil metal particles. Based on the wind turbine's operational observation data, thermal images and external images of the blades are acquired. The thermal images are processed to calculate the thermal transfer data. Based on the thermal transfer data, a thermal transfer standard is obtained. The external images of the blades are processed to obtain the grayscale values. Based on the grayscale values, a grayscale value standard is obtained. The thermal transfer standard and the grayscale value standard constitute the observation standard. The blade rotation speed information is processed as follows: Based on the blade rotation speed information, the blade rotation speed ypz(h) is obtained. The blade rotation speeds of different wind turbines are obtained and denoted as ypz(h, c). The blade speeds ypz(h,c) of wind turbines under the same environmental influence are statistically analyzed, and the maximum and minimum speeds from ypz(h,1) to ypz(h,cs) are extracted; the maximum speed is denoted as zdz(h) and the minimum speed is denoted as zxz(h). The values ​​of environmental parameter h are obtained and denoted as h1 to hs; the maximum and minimum speeds of the fan under different environmental parameters are statistically analyzed to obtain zdz(h1) to zdz(hs) and zxz(h1) to zxz(hs). Based on the maximum and minimum speeds of the fan under different environmental parameters, a speed environment model hjm is constructed. ； Where mys(h) represents the preset speed, which refers to the actual blade speed under environmental parameter h; The real-time speed of the fan is obtained, and the real-time speed is substituted into the speed environment model for calculation to analyze speed anomalies. Data analysis module: performs real-time monitoring of the wind turbine, acquires monitoring data, judges the operating status and observation standards of the wind turbine based on data thresholds, analyzes abnormal wind turbine operation information, and analyzes abnormal wind turbine structure information based on monitoring data and observation standards. Fault warning module: Provides fault warnings based on abnormal operation and structural abnormalities of the wind turbine, and performs intelligent troubleshooting of wind turbine faults.

2. The intelligent troubleshooting system for wind turbine operation faults according to claim 1, characterized in that, The data acquisition module collects data, as follows: The angular velocity of the blade is obtained, and the rotational speed of the blade is obtained based on the angular velocity of the blade. The rotational speed of the blade under different environments is obtained to obtain the blade rotational speed ypz(h). The blade rotation speed is statistically analyzed to obtain the blade rotation speed information; Set up as measurement points on the blade. Based on the number of measurement points, obtain the distance between each measurement point and the centerline of the blade to get the measurement distance clj(a). Statistically analyze the measurement distances to obtain the vibration data of the blade. The number of oil metal particles in the fan is collected, and the metal particle category bs is obtained. According to the metal particle category bs, the number of each type of oil metal particle is recorded as yjk(b). The number of oil metal particles is counted to obtain the wear information of the fan. The operating status data of the wind turbine is composed of the blade rotation speed information, blade vibration data and wind turbine wear information; Images of the blade surface are acquired to obtain the appearance information of the blade, the thermal distribution of the blade is obtained, and a thermal image of the blade is constructed based on the thermal distribution of the blade. The wind turbine's operational observation data are constructed based on the thermal images and external images of the blades.

3. The intelligent troubleshooting system for wind turbine operation faults according to claim 1, characterized in that, The vibration data of the blades is processed as follows: Based on the vibration data of the blade, the measured distance at different angles when the blade rotates one revolution is obtained. The measured distance is discretized and ds different angles are extracted. Based on the ds different angles, the measured distance is denoted as clj(a,d). Under static conditions, the distance between the measurement point and the center of the blade is obtained, which is the static distance jzj(a). The vibration angle of the blade is obtained by calculating the measurement distance clj(a,d) based on the static distance jzj(a). The vibration threshold zyz(a) is obtained by calculating the mean of the vibration angle of the blade.

4. The intelligent troubleshooting system for wind turbine operation faults according to claim 1, characterized in that, The wear information of the wind turbine is processed as follows: Based on the wear information of the fan, the number of oil metal particles yjk(b,c) of different fans is obtained. Based on the number of oil metal particles yjk(b,c), the number of oil metal particles for each type of oil is counted, and the maximum value of the number of oil metal particles is extracted to obtain the maximum number of particles zdk(b). The maximum number of particles zdk(b) is used as the single-class threshold. The total number of oil-metal particles in each fan is obtained by counting all types of oil-metal particles in the fan. The maximum value of the total number of oil-metal particles is obtained to get the total threshold zsy. The wear threshold of the wind turbine is composed of the single-category threshold and the total number threshold.

5. The intelligent troubleshooting system for wind turbine operation faults according to claim 1, characterized in that, The thermal and external images of the blades were processed as follows: Based on the thermal image of the blade, the number of pixels in the thermal image is obtained and denoted as is×js; the temperature of each pixel is obtained and the pixel temperature xwd(i,j) is obtained. The pixel temperature xwd(i,j) is iterated and calculated to obtain the temperature transfer amounts cd1(i,j) to cd4(i,j) in different directions. Based on the temperature transfer amounts cd1(i,j) to cd4(i,j) in different directions, the temperature transfer amount of all pixels on the thermal image is statistically calculated to obtain the thermal transfer standard cdb. Based on the appearance image of the blade, the grayscale value of the image is obtained, and the appearance image is aligned with the thermal image to make the number of pixels in the appearance image is×js. Based on the number of pixels, the grayscale value of each pixel position is obtained to get the pixel grayscale value hdz(i,j); based on the pixel grayscale value, the grayscale values ​​of each row of the blade appearance image are traversed to get the row grayscale value hhd(i). The column grayscale value lhd(j) is obtained by iterating through each column of grayscale values ​​in the appearance image of the blade based on the pixel grayscale values. The grayscale values ​​hhd(i) and lhd(j) of the rows and columns are statistically analyzed to obtain the grayscale value standard.

6. The intelligent troubleshooting system for wind turbine operation faults according to claim 1, characterized in that, The operating status and observation standards of the wind turbines are judged as follows: The fan is monitored in real time to obtain monitoring data. Based on the monitoring data, the real-time speed of the fan is obtained. The real-time speed is substituted into the speed environment model for calculation, and speed anomalies are analyzed. The real-time distance of the measurement point is obtained. Based on the real-time distance of the measurement point, the vibration angle of the fan blades is calculated. Combined with the vibration threshold, vibration anomalies are analyzed. The real-time number of metal particles in the fan oil is obtained. Combined with the fan wear threshold, wear anomalies are analyzed. Based on the fan speed anomalies, vibration anomalies, and wear anomalies, the fan's operational anomaly information is constructed. Based on the monitoring data of the wind turbine, real-time monitoring images of the wind turbine are acquired. Real-time thermal images and real-time appearance images of the wind turbine are obtained through these images. The real-time thermal images of the wind turbine are processed in conjunction with heat transfer standards to analyze heat transfer anomalies. The real-time appearance images of the wind turbine are processed in conjunction with grayscale standards to analyze appearance anomalies. The structural anomalies of the wind turbine are constituted by the heat transfer anomalies and appearance anomalies.

7. The intelligent troubleshooting system for wind turbine operation faults according to claim 6, characterized in that, Information on abnormal operation of the wind turbine is collected, as follows: The real-time speed and speed environment model of the wind turbine are obtained. The real-time speed is substituted into the speed environment model to obtain the speed judgment value zpd. The speed judgment value zpd is then analyzed. If the speed judgment value zpd is greater than 0, it indicates that the fan speed is abnormal; if the speed judgment value zpd is less than or equal to 0, it indicates that the fan speed is normal. The real-time distance of the measurement point is obtained and denoted as ssj(a). The stationary distance jzj(a) is obtained. Based on the real-time distance and the stationary distance, the real-time vibration angle is calculated. Combined with the vibration threshold, the vibration judgment value zpd(a) is obtained. ； If the vibration judgment value zpd(a) is greater than 0, it indicates that the fan has abnormal vibration, and the abnormal location is the position of the a-th measurement point; if the vibration judgment value zpd(a) is less than or equal to 0, it indicates that the fan vibration is normal. The real-time number of oil metal particles in the fan is obtained and denoted as syl(b). Combined with the wear threshold of the fan, the total wear judgment value mpd and the single metal wear judgment value mpd(b) are obtained. ； Where mpd is the judgment value of the total amount of all worn metal particles, mpd(b) is the wear judgment of the metal particles in the bth class; zsy is the total threshold, and zdk(b) is the single class threshold; If mpd is greater than 0, it indicates abnormal wear of the fan. If mpd(b) is greater than 0, it indicates abnormal wear of the b-th type of metal particles in the fan. If mpd is less than or equal to 0 and mpd(b) is less than or equal to 0, it indicates normal wear of the fan.

8. The intelligent troubleshooting system for wind turbine operation faults according to claim 6, characterized in that, The analysis of abnormal heat transfer and appearance of the fan is as follows: The real-time thermal image of the fan is acquired, and the real-time temperature at different pixel positions is obtained based on the real-time thermal image of the fan, denoted as swd(i,j); the real-time temperature is traversed, and the heat transfer judgment value rcp(i,j) is calculated in combination with the heat transfer standard cdb. ； If the heat transfer judgment value rcp(i,j) is greater than 0, it indicates that the heat transfer of the fan is abnormal, and the abnormal position is the position corresponding to the image pixel (i,j); if the heat transfer judgment value rcp(i,j) is less than 0, it indicates that the heat transfer of the fan is normal. Acquire real-time appearance images of the wind turbine, obtain the grayscale value of each pixel in the real-time appearance image of the wind turbine, and obtain the real-time grayscale value shd(i,j); calculate the grayscale judgment value based on the real-time grayscale value and the grayscale value standard. ； Where hdp(i) represents the grayscale value judgment value for each row of pixels, ldp(j) represents the grayscale value judgment value for each column of pixels; hhd(i) and lhd(j) are grayscale value standards; When hdp(i) is not equal to 0, it indicates that there is an anomaly in the i-th row of pixels. When ldp(j) is not equal to 0, it indicates that there is an anomaly in the j-th column of pixels. Based on the statistics of hdp(i) and ldp(j), the appearance anomaly location is obtained.

Images