CN118148852A - Fan blade fault detection method, device, equipment and readable storage medium - Google Patents

Abstract

The application discloses a fan blade fault detection method, a device, equipment and a readable storage medium, which are applied to the technical field of fan blade detection and comprise the following steps: acquiring the vibration acceleration of the blade, and determining the actual vibration mode frequency of the blade by utilizing the vibration acceleration of the blade to perform fast Fourier transformation; collecting axial vibration acceleration of the main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out fast Fourier transformation to determine axial actual rotation frequency of the main shaft and axial actual rotation frequency amplitude of the main shaft; and determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the actual spindle axial rotation frequency and the amplitude thereof. According to the application, the fan blade faults are determined by utilizing the blade modal frequency characteristics and the unbalanced vibration characteristics of the wind wheel, which are directly and closely related to the fan blade, and because the abnormal judgment indexes of the blade vibration acceleration and the main shaft axial vibration acceleration are obtained through fault mechanism analysis, the indexes which do not depend on a large amount of historical data can realize the accurate diagnosis of the wind turbine blade faults.

Description

Fan blade fault detection method, device, equipment and readable storage medium

Technical Field

The invention relates to the technical field of fan blade detection and blade state monitoring, in particular to a fan blade fault detection method, device and equipment and a readable storage medium.

Background

The wind turbine blade is used as a key component for energy conversion of the wind turbine, and the reliability of the blade plays a vital role in generating efficiency and safe operation of the wind turbine. At present, no mature and effective blade online monitoring technology exists in the field, in recent years, a plurality of detection methods are proposed by students and experts, online monitoring is difficult to realize, and in addition, the students propose to monitor blade pitch angle and power by using a SCADA (supervisory control and data acquisition) system, and the blade condition is detected by performing deviation analysis on monitored values and theoretical values of the monitored blade pitch angle and power and setting corresponding pre-alarm thresholds. However, as the working state of the wind turbine generator is complex, the working condition is influenced by various factors, the variable pitch angle of the blades and the power dispersion are large, and therefore, the theoretical value of the physical quantity is difficult to calculate accurately, and the method is low in pertinence and accuracy.

Therefore, the existing method for detecting the faults of the fan blades has the technical problem of low accuracy.

Disclosure of Invention

Accordingly, the present invention is directed to a fan blade fault detection method, apparatus, device and readable storage medium, which solve the technical problem of low fan blade fault detection accuracy in the prior art.

In order to solve the technical problems, the invention provides a fan blade fault detection method, which comprises the following steps:

collecting the vibration acceleration of the blade, and performing fast Fourier transformation by using the vibration acceleration of the blade to determine the actual vibration mode frequency of the blade;

collecting axial vibration acceleration of a main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out the fast Fourier transformation to determine the axial actual rotation frequency of the main shaft and the axial actual rotation frequency amplitude of the main shaft;

And determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof.

Optionally, the collecting the vibration acceleration of the blade, and performing fast fourier transform to determine the actual vibration mode frequency of the blade by using the vibration acceleration of the blade includes:

Performing integral operation on the blade vibration acceleration to determine blade vibration displacement;

After the fast fourier transform operation is performed on the blade vibration displacement, determining a first preset search range by utilizing theoretical blade vibration modal frequency, searching the maximum spectrum amplitude in the first preset search range, and determining the blade vibration modal frequency corresponding to the maximum spectrum amplitude as the actual blade vibration modal frequency.

Optionally, the determining whether the fan blade has a fault by using the actual blade vibration modal frequency, the spindle axial actual rotation frequency and the amplitude thereof includes:

Dividing the difference value of the actual blade vibration mode frequency and the theoretical blade vibration mode frequency by the theoretical blade mode frequency to obtain blade mode frequency deviation;

and determining whether the blade modal frequency deviation exists or not by using the blade modal frequency deviation and the modal frequency deviation threshold value.

Optionally, the collecting the spindle axial vibration acceleration, and performing the fast fourier transform by using the spindle axial vibration acceleration to determine a spindle axial actual rotation frequency and a spindle axial actual rotation frequency amplitude, including:

Performing integral operation on the axial vibration acceleration of the spindle to obtain axial vibration displacement of the spindle;

And carrying out the fast Fourier transform operation on the axial vibration displacement of the spindle, determining a second preset searching range by utilizing the average frequency conversion of the wind wheel, searching the maximum frequency spectrum amplitude in the second preset searching range, and determining the spindle axial frequency conversion corresponding to the maximum frequency spectrum amplitude and the amplitude thereof as the spindle axial actual frequency conversion and the spindle axial actual frequency conversion amplitude.

Optionally, the determining whether the fan blade has a fault by using the actual blade vibration modal frequency, the spindle axial actual rotation frequency and the amplitude thereof includes:

Determining a blade vibration reference frequency by using the product of the actual axial rotation frequency of the spindle and the number of blades, determining a third preset search range by using the blade vibration reference frequency, searching the maximum frequency spectrum amplitude in the third preset search range, and determining the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the spindle;

Determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the actual frequency conversion amplitude of the axial direction of the main shaft to the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft;

And determining whether the unbalanced vibration condition of the wind wheel is met according to the frequency conversion-blade frequency amplitude ratio and the frequency conversion-blade frequency amplitude threshold value.

Optionally, the fan blade fault detection method further includes:

Determining whether the blade modal frequency deviation exceeds a threshold, whether the frequency-to-blade frequency amplitude exceeds the threshold and whether abrupt changes exist in a trend curve of the frequency-to-blade frequency amplitude ratio under the blade icing meteorological conditions when the blade icing meteorological conditions are met;

And if the blade mode frequency deviation exceeds a threshold, the frequency conversion-blade frequency amplitude ratio exceeds the threshold, and a mutation exists in a trend curve of the frequency conversion-blade frequency amplitude ratio, determining that the blade icing fault exists.

Optionally, before the determining that the blade icing meteorological conditions are met, the method further comprises:

Acquiring meteorological factors in a blade working environment; wherein the meteorological factors include air temperature, air humidity and air pressure;

Determining whether the air temperature is less than or equal to the preset temperature threshold, whether the air humidity is greater than or equal to the preset humidity threshold, and whether the air pressure is less than or equal to the preset air pressure threshold according to a preset temperature threshold, a preset humidity threshold and a preset air pressure threshold;

And when the air temperature is determined to be less than or equal to the preset temperature threshold value, the air humidity is determined to be greater than or equal to the preset humidity threshold value, and the air pressure is determined to be less than or equal to the preset air pressure threshold value, determining that the blade icing meteorological conditions are met.

Optionally, the determining process for the mutation in the trend curve of the frequency conversion-leaf frequency amplitude ratio includes:

determining trend mutation factors corresponding to the two time points; wherein the formula of the trend mutation factor is Wherein i and k respectively represent different time points, w ₂ (i) represents the frequency conversion-leaf frequency amplitude ratio of the i point, and w ₂ (k) represents the frequency conversion-leaf frequency amplitude ratio of the k point;

determining whether the trend mutation factor is greater than a trend mutation threshold value;

when greater than, the presence of a mutation is determined.

Optionally, the determining whether the fan blade has a fault by using the actual blade vibration modal frequency, the spindle axial actual rotation frequency and the amplitude thereof includes:

Determining a blade mode frequency deviation by utilizing the ratio of the difference between the actual blade vibration mode frequency and the theoretical blade mode frequency to the theoretical blade mode frequency, and taking the blade mode frequency deviation as a first type of characteristic quantity;

Determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the spindle axial actual frequency conversion amplitude of the spindle axial actual frequency conversion to the maximum frequency spectrum amplitude of the spindle axial vibration displacement frequency spectrum, and taking the frequency conversion-leaf frequency amplitude ratio as a second type of characteristic quantity;

constructing a first feature matrix by utilizing a plurality of first type feature quantities and a plurality of second type feature quantities corresponding to a plurality of continuous samples; wherein the elements in the first feature matrix represent the values of the nth class feature quantity of the mth sample;

performing super-threshold calculation on the first feature matrix to obtain a second feature matrix;

And calculating the weight of each characteristic quantity in the second characteristic matrix by adopting an entropy method, and realizing comprehensive decision on the fault state of the blade based on a linear weighting model and each characteristic threshold value.

Optionally, the calculating the weight of each feature in the second feature matrix by adopting an entropy method, and implementing the comprehensive decision on the fault state of the blade based on the linear weighting model and each feature threshold value includes:

Normalizing the values of all the feature quantities of each sample of the second feature matrix;

calculating entropy values of all the characteristic quantities by using an entropy method, and determining deviation degree by using the entropy values;

normalizing the deviation degree to obtain weight values of various characteristic quantities,

Constructing the weight value to obtain a decision matrix;

And carrying out comprehensive decision on each weight value in the decision matrix by utilizing the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and the corresponding comprehensive decision limiting threshold value.

Optionally, the comprehensively deciding each weight value in the decision matrix by using the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and a corresponding comprehensive decision limiting threshold value, including:

And carrying out comprehensive decision on each weight value by using the linear weighting model to determine a comprehensive decision value, wherein the linear weighting model is as follows: y _i is the integrated decision value of the ith sample, x _ij is the value of the jth feature quantity of the ith sample, and w _j is the weight value of the jth feature.

Optionally, acquiring the blade vibration acceleration through a sensor on the fan blade; and collecting axial vibration acceleration of the main shaft through a vibration sensor below the main shaft of the fan.

The application also provides a fan blade fault detection device, which comprises:

The actual blade vibration mode frequency determining module is used for collecting the blade vibration acceleration and performing fast Fourier transformation by utilizing the blade vibration acceleration to determine the actual blade vibration mode frequency;

the main shaft axial actual rotation frequency determining module is used for collecting main shaft axial vibration acceleration and utilizing the main shaft axial vibration acceleration to carry out the fast Fourier transformation to determine main shaft axial actual rotation frequency and main shaft axial actual rotation frequency amplitude;

And the blade fault determining module is used for determining whether the fan blade has faults or not by utilizing the actual blade vibration modal frequency, the spindle axial actual rotation frequency and the amplitude thereof.

The application also provides fan blade fault detection equipment, which comprises:

A memory for storing a computer program;

And the processor is used for realizing the fan blade fault detection method when executing the computer program.

The application also provides a readable storage medium, wherein the readable storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by a processor, the fan blade fault detection method is realized.

Therefore, the application determines the actual blade vibration mode frequency by acquiring the blade vibration acceleration in a preset time period and performing fast Fourier transformation by utilizing the blade vibration acceleration; collecting spindle axial vibration acceleration in a preset time period, and utilizing the spindle axial vibration acceleration to perform the fast Fourier transformation to determine spindle axial actual rotation frequency and amplitude; and determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the actual spindle axial rotation frequency and the amplitude thereof. Compared with the prior art, the application provides a method for realizing the identification of the wind turbine blade faults based on the multi-vibration characteristic joint diagnosis, firstly, the two vibration characteristics of the blade vibration acceleration of the wind turbine blade and the axial vibration acceleration of the main shaft are extracted to identify the wind turbine blade faults, and because the abnormal judgment indexes of the blade vibration acceleration and the axial vibration acceleration of the main shaft are indexes which are obtained through fault mechanism analysis and do not depend on a large amount of historical data, the accurate diagnosis of the wind turbine blade faults can be realized; the quantitative analysis and the accurate diagnosis of the fault reliability are further realized by adopting a multi-sample and multi-feature entropy weighting comprehensive decision method, so that the error decision caused by an abnormal single sample is avoided, and the accurate diagnosis of the wind turbine blade fault is realized as a whole; in addition, a specific method for judging the icing of the blade through vibration data indexes is provided, and the icing fault of the blade can be further identified on the aspects of weather factor characteristics such as combined temperature and the like; in addition, the detection system can realize the real-time monitoring of fan blade faults, is convenient to install, has simple data processing flow, is easy to realize, has good diagnosis effect and has good engineering application value.

In addition, the invention also provides a fan blade fault detection device, equipment and a readable storage medium, which have the same beneficial effects.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a fan blade failure detection method provided by an embodiment of the present invention;

FIG. 2 is a flowchart illustrating a method for extracting a blade mode frequency feature according to an embodiment of the present invention;

fig. 3 is a flowchart illustrating a method for extracting unbalanced vibration characteristics of a wind wheel according to an embodiment of the present invention;

FIG. 4 is a flowchart illustrating a method for identifying a blade icing fault according to an embodiment of the present invention;

FIG. 5 is a flowchart illustrating a multi-feature comprehensive decision method according to an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a fan blade failure detection device according to an embodiment of the present invention;

Fig. 7 is a schematic structural diagram of a fan blade fault detection device according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Referring to fig. 1, fig. 1 is a flowchart of a fan blade fault detection method according to an embodiment of the present invention. The method may include:

s101, acquiring the vibration acceleration of the blade, and determining the actual vibration mode frequency of the blade by utilizing the vibration acceleration of the blade to perform fast Fourier transformation.

The execution body of this embodiment is any electronic device, such as a mobile phone, a computer, a specially designed fault monitoring and diagnosing instrument, and the like. In the embodiment, the axial vibration acceleration of the main shaft can be acquired by additionally arranging a vibration sensor below the main shaft of the wind turbine. It can be appreciated that in order to ensure the safety of the operation condition of the wind turbine, the method can be utilized to monitor the wind turbine in real time. This embodiment acquires the blade vibration acceleration by the sensor. It will be appreciated that the wind turbine blade changes its own mass or stiffness due to icing or structural damage, and that the change in the blade's own mass or stiffness results in a consequent change in the actual blade vibration mode frequency. Therefore, the modal frequency may be used as a feature to monitor the blade condition.

It should be further noted that, the above-mentioned collection of the blade vibration acceleration and the rapid fourier transform of the blade vibration acceleration are used to determine the actual blade vibration mode frequency, which may include: performing integral operation on the vibration acceleration of the blade to determine the vibration displacement of the blade; after the fast fourier transform operation is performed on the blade vibration displacement, a first preset search range is determined by utilizing the theoretical blade vibration modal frequency, the maximum frequency spectrum amplitude is searched in the first preset search range, and the blade vibration modal frequency corresponding to the maximum frequency spectrum amplitude is determined as the actual blade vibration modal frequency. The theoretical blade vibration mode frequency in this embodiment is a design value, or may be referred to as a fixed value, which is a value that the wind turbine can develop when designed according to design parameters. In this embodiment, a first preset search range is determined by using the theoretical blade vibration mode frequency, a maximum spectrum amplitude is searched in the first preset search range, and the blade vibration mode frequency corresponding to the maximum spectrum amplitude is determined as the actual blade vibration mode frequency.

S102, collecting the axial vibration acceleration of the main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out fast Fourier transformation to determine the axial actual rotation frequency of the main shaft and the axial actual rotation frequency amplitude of the main shaft.

According to the embodiment, the vibration sensor is additionally arranged below the main shaft of the wind turbine, so that the axial vibration acceleration of the main shaft can be acquired within a period of time. According to the embodiment, the axial vibration acceleration of the main shaft is utilized to carry out integral operation, so that axial vibration displacement of the main shaft is obtained, wind wheel rotating speed data of a SCADA system (data acquisition and monitoring control system) in the same time period are obtained, and the wind wheel average rotating frequency corresponding to the wind wheel average rotating speed in the time period is calculated. More preferably, because the rotating speed of the wind wheel is low, the influence of the rotating speed fluctuation on the rotating frequency and the passing frequency of the blades is large, and if the real-time rotating speed signal of the wind wheel can be obtained, a more accurate rotating frequency value in the monitoring period can be calculated. And (3) carrying out FFT (fast Fourier transform) operation on axial vibration displacement of the spindle, according to the step of searching the vibration modal frequency of the actual blade, calculating to obtain the axial actual rotation frequency of the spindle according to the average rotation frequency of the wind wheel, namely setting delta f as an actual rotation frequency searching frequency band f _b＝[f_n0-Δf,f_n0 plus delta f from left to right by using the center of the average rotation frequency f _n0 of the wind wheel, and searching the maximum frequency spectrum amplitude of the displacement frequency spectrum s (f _b), wherein the frequency corresponding to the maximum frequency amplitude is the axial actual rotation frequency f _nx of the spindle. It will be appreciated that there is a difference in the mass or aerodynamic properties of each blade, resulting in a non-smooth rotation of the rotor as a whole, i.e. a rotor imbalance failure. When the fan normally operates, only obvious blade vibration frequency exists in the axial vibration signal frequency spectrum of the wind wheel output shaft, and the blade vibration frequency is multiple times of that of the wind wheel rotating frequency. If unbalanced vibration occurs to the wind wheel, the axial vibration of the main shaft additionally generates frequency conversion, which is the reason why the rotation of the wind wheel is modulated, and the more serious the unbalance of the wind wheel is, the more obvious the frequency conversion component is. Therefore, the spindle axial rotation 1-order vibration (spindle axial actual rotation) can be used as a characteristic quantity of the blade state.

S103, determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof.

The embodiment is not limited to a specific manner of determining whether a fan blade has a fault using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof. For example, the actual blade vibration mode frequency and the spindle axial actual rotation frequency can be compared with the corresponding fault thresholds, if both the actual blade vibration mode frequency and the spindle axial actual rotation frequency exceed the fault thresholds, the fan blade is determined to have faults, otherwise, the fan blade is determined to have no faults; or the embodiment can determine the blade mode frequency deviation according to the difference value between the actual blade vibration mode frequency and the blade mode frequency reference value and the ratio of the blade mode frequency reference value, compare the blade mode frequency deviation with a threshold value, determine whether the deviation exists (whether the deviation is normal), determine the blade vibration reference frequency according to the product of the number of wind turbine blades and the axial actual rotation frequency of the main shaft, obtain the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft according to the actual frequency searching method, and determine the frequency corresponding to the axial actual rotation frequency of the main shaft. The ratio of the actual axial rotation frequency amplitude of the main shaft to the maximum frequency amplitude of the axial vibration displacement frequency spectrum of the main shaft is determined to be the rotation frequency-blade frequency amplitude ratio in the axial vibration displacement frequency spectrum, whether the rotation frequency-blade frequency amplitude ratio in the axial vibration displacement frequency spectrum is larger than a corresponding threshold value is determined, when the rotation frequency-blade frequency amplitude ratio is larger than the corresponding threshold value, the existence of unbalanced wind wheel is determined, namely the existence of deviation of modal frequency deviation of the blade is satisfied, and when the unbalanced wind wheel is satisfied, the existence of faults of the fan blade is determined; or the embodiment can further determine the blade mode frequency deviation corresponding to a plurality of times and the conversion-blade frequency amplitude ratio in the axial vibration displacement frequency spectrum according to a plurality of samples, so as to construct a first feature matrix, wherein elements in the feature matrix represent blade mode frequency deviation feature values of the nth sample and the conversion-blade frequency amplitude ratio in the axial vibration displacement frequency spectrum of the nth sample, threshold feature matrices corresponding to the threshold feature matrices are constructed by utilizing threshold values corresponding to the conversion-blade frequency amplitude ratios in each blade mode frequency deviation and the axial vibration displacement frequency spectrum, the first feature matrix is differed from the threshold feature matrices, a second feature quantity matrix is obtained, weights of various feature values in the second feature matrix are determined based on an entropy method, comprehensive determination is performed by adopting a linear weighting model based on the weights of various feature values, and parameter values corresponding to each sample are obtained, so that whether a fan blade has faults or not is determined according to the set threshold value.

It should be further noted that, to improve the accuracy of determining the fan fault according to the actual blade vibration mode frequency, determining whether the fan blade has a fault by using the actual blade vibration mode frequency and the actual spindle axial rotation frequency and the amplitude thereof may include: dividing the difference value of the actual blade vibration mode frequency and the theoretical blade vibration mode frequency by the theoretical blade mode frequency to obtain blade mode frequency deviation; and determining whether the blade modal frequency deviation exists or not by using the blade modal frequency deviation and the modal frequency deviation threshold value. The embodiment can determine whether the blade mode frequency deviation exists or not according to the blade mode frequency deviation and the mode frequency deviation threshold value, and the blade mode frequency deviation can be used as one index for determining whether the fan blade has faults or not.

It should be further noted that, in order to improve the accuracy of determining the spindle axial actual rotation frequency and the amplitude thereof, the collecting the spindle axial vibration acceleration, and performing the fast fourier transform by using the spindle axial vibration acceleration to determine the spindle axial actual rotation frequency and the spindle axial actual rotation frequency amplitude may include: performing integral operation on the axial vibration acceleration of the main shaft to obtain axial vibration displacement of the main shaft; and performing fast Fourier transform operation on the axial vibration displacement of the spindle, determining a second preset search range by utilizing the average rotation frequency of the wind wheel, searching the maximum frequency spectrum amplitude in the second preset search range, and determining the spindle axial rotation frequency and the amplitude thereof corresponding to the maximum frequency spectrum amplitude as the spindle axial actual rotation frequency and the spindle axial actual rotation frequency amplitude. The method of obtaining the average rotor frequency of the wind wheel is not limited in this embodiment. The embodiment can acquire the wind wheel rotating speed data v (t) of a SCADA (data acquisition and monitoring control system) system in the same period t, and calculate the average wind wheel rotating speed corresponding to the wind wheel rotating speed v (t) in the periodCorresponding wind wheel average rotation frequency f _n0,/>Or better, because the rotating speed of the wind wheel is low, the influence of the rotating speed fluctuation on the rotating frequency and the passing frequency of the blades is larger, and if the real-time rotating speed signal of the wind wheel can be obtained, a more accurate rotating frequency value in the monitoring period can be calculated. The embodiment provides a specific method for determining the spindle axial actual rotation frequency and the spindle axial actual rotation frequency amplitude, and improves the accuracy of the spindle axial actual rotation frequency and the amplitude determination thereof.

It should be further noted that, determining whether the fan blade has a fault by using the actual blade vibration mode frequency and the spindle axial actual rotation frequency may include: determining a blade vibration reference frequency by using the product of the actual axial rotation frequency of the spindle and the number of blades, determining a third preset search range by using the blade vibration reference frequency, searching the maximum frequency spectrum amplitude in the third preset search range, and determining the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the spindle; determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the actual frequency conversion amplitude of the axial direction of the main shaft to the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft; and determining whether the unbalanced vibration condition of the wind wheel is met according to the frequency conversion-blade frequency amplitude ratio and the frequency conversion-blade frequency amplitude threshold value. According to the embodiment, the blade vibration reference frequency is determined according to the product of the number of wind turbine blades and the actual axial rotation frequency of the main shaft, and in order to avoid error in blade frequency searching caused by sampling precision, a third preset searching range is determined by utilizing the blade vibration reference frequency according to the actual frequency searching method, and the maximum frequency spectrum amplitude is searched in the third preset searching range, so that the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft is obtained. The ratio of the axial actual rotation frequency amplitude of the main shaft to the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft is used for determining the rotation frequency-blade frequency amplitude ratio, and whether unbalanced vibration of the wind wheel exists or not is determined by setting a related limiting threshold value, and when the limiting threshold value is exceeded, the unbalanced vibration condition of the wind wheel is determined to be met. And if the modal frequency of the blade and the unbalanced vibration characteristic value of the wind wheel exceed the limiting threshold, judging that the blade fault exists. The embodiment provides a specific method for specifically calculating the axial actual rotation frequency of the main shaft and determining the unbalance of the wind wheel, and improves the accuracy of determining the unbalance of the wind wheel.

It should be further noted that, in order to implement the method for identifying the blade icing fault, the method for detecting the fan blade fault may further include: when the weather conditions of the blade icing are met, judging whether the modal frequency deviation of the blade exceeds a threshold, whether the amplitude of the frequency conversion and the blade frequency exceeds the threshold and whether abrupt changes exist in a trend curve of the amplitude ratio of the frequency conversion and the blade frequency; if the blade mode frequency deviation exceeds the threshold, the frequency conversion-blade frequency amplitude ratio exceeds the threshold, and a mutation exists in a trend curve of the frequency conversion-blade frequency amplitude ratio, determining that the blade icing fault exists. Based on the blade vibration characteristic results, the embodiment combines wind field meteorological factors to judge whether the blade icing fault exists. Weather factors including, but not limited to, air temperature, air humidity, and air pressure in the blade operating environment are obtained and corresponding temperature, humidity, and air pressure thresholds are set. The following 3 conditions are satisfied: ① The air temperature is less than or equal to a preset temperature threshold; ② The air humidity is greater than or equal to a preset humidity threshold; ③ The air pressure is smaller than or equal to a preset air pressure threshold value, and meets the weather conditions of blade icing. If the modal frequency of the blade and the unbalance of the wind wheel are both over-limited only under the blade icing meteorological condition, and the trend curve of the frequency conversion-blade frequency amplitude ratio has mutation, judging that the blade unbalance fault exists, otherwise, the blade unbalance fault belongs to the self fault of the blade and the non-icing fault. The embodiment does not limit the specific determination of the presence of a sudden change in the trend curve of the frequency-to-leaf frequency amplitude ratio. For example, whether there is a mutation may be determined based on whether the difference of the consecutive two of the frequency-to-leaf frequency amplitudes is greater than a preset threshold, or whether there is a mutation may be determined based on the fluctuation amplitude between the differences of the consecutive two of the frequency-to-leaf frequency amplitudes being such that it is greater than a threshold amplitude value; or the trend mutation factor can be determined according to the amplitude ratio of the frequency conversion and the leaf frequency, and whether mutation exists or not can be determined according to the comparison between the trend mutation factor and the mutation threshold value. According to the embodiment, the specific steps of fan blade fault judgment under the blade icing condition are provided, so that the accuracy of fan blade fault determination is improved.

It should be further noted that, in order to improve the accuracy of the blade icing fault determination, before the determination meets the blade icing meteorological condition, the method may further include: acquiring meteorological factors in a blade working environment; wherein, the meteorological factors comprise air temperature, air humidity and air pressure; determining whether the air temperature is less than or equal to the preset temperature threshold, whether the air humidity is greater than or equal to the preset humidity threshold, and whether the air pressure is less than or equal to the preset air pressure threshold according to the preset temperature threshold, the preset humidity threshold and the preset air pressure threshold; and when the air temperature is determined to be smaller than or equal to a preset temperature threshold value, the air humidity is determined to be larger than or equal to a preset humidity threshold value, and the air pressure is determined to be smaller than or equal to a preset air pressure threshold value, determining that the blade icing meteorological conditions are met. The embodiment provides three specific conditions for judging the blade icing meteorological conditions, and improves the accuracy of blade icing meteorological conditions determination.

It should be further noted that, in order to improve accuracy of determining the mutation in the trend curve of the frequency-to-leaf frequency amplitude ratio, the determining process of the mutation in the trend curve of the frequency-to-leaf frequency amplitude ratio may include: determining trend mutation factors corresponding to the two time points; wherein, the formula of the trend mutation factor isWherein i and k respectively represent different time points, w ₂ (i) represents the frequency conversion-leaf frequency amplitude ratio of the i point, and w ₂ (k) represents the frequency conversion-leaf frequency amplitude ratio of the k point; determining whether the trend mutation factor is greater than a trend mutation threshold value; when greater than, the presence of a mutation is determined. According to the embodiment, the trend mutation factor is determined according to the frequency conversion-leaf frequency amplitude ratio of the two time points, so that whether mutation exists is determined directly according to the trend mutation factor and the trend mutation threshold value, specific parameters for judging mutation are given, and the accuracy of determining mutation in a trend curve of the frequency conversion-leaf frequency amplitude ratio is improved.

It should be further noted that, to improve the accuracy of determining the failure of the fan blade, determining whether the fan blade has a failure by using the actual vibration mode frequency of the fan blade and the actual rotation frequency of the spindle axis and the amplitude thereof may include:

s1031, determining the blade mode frequency deviation by utilizing the ratio of the difference between the actual blade vibration mode frequency and the theoretical blade mode frequency to the theoretical blade mode frequency, and taking the blade mode frequency deviation as a first type of characteristic quantity;

S1032, determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the spindle axial actual frequency conversion amplitude of the spindle axial actual frequency conversion to the maximum frequency spectrum amplitude of the spindle axial vibration displacement frequency spectrum, and taking the frequency conversion-leaf frequency amplitude ratio as a second type of characteristic quantity;

S1033, constructing a first feature matrix by utilizing a plurality of first type feature quantities and a plurality of second type feature quantities corresponding to the continuous plurality of samples; wherein the elements in the first feature matrix represent the values of the nth class feature values of the mth sample;

s1034, performing super-threshold calculation on the first feature matrix to obtain a second feature matrix;

S1035, calculating the weight of each feature quantity in the second feature matrix by adopting an entropy method, and realizing comprehensive decision on the fault state of the blade based on the linear weighting model and each feature threshold value.

According to the embodiment, after the actual blade vibration mode frequency, the actual spindle axial rotation frequency and the amplitude thereof are obtained, a feature matrix is also constructed, so that the detection of the blade fault state is realized according to the feature matrix. The embodiment can realize comprehensive decision on the fault state of the blade by utilizing an entropy method and linear weighting, and improves the accuracy of determining the fault of the fan blade.

It should be further noted that, in order to improve accuracy of determining the fan blade fault by using the feature matrix, the calculating the weight of each feature quantity in the second feature matrix by using the entropy method, and implementing the comprehensive decision on the blade fault state based on the linear weighting model and each feature threshold may include:

normalizing the values of all the characteristic quantities of each sample of the second characteristic matrix;

calculating entropy values of all the characteristic quantities by using an entropy method, and determining deviation degree by using the entropy values;

Normalizing the deviation degree to obtain weight values of various characteristic quantities,

Constructing the weight value to obtain a decision matrix;

And comprehensively deciding each weight value in the decision matrix by utilizing the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and the corresponding comprehensive decision limiting threshold value.

The comprehensive decision limiting threshold value in the embodiment can be multiple, and for example, the comprehensive decision limiting threshold value can comprise an early warning value and an alarm value; or may include a warning value, an early warning value, and an alarm value. The number of the comprehensive decision limiting threshold values in the embodiment is multiple, so that the comprehensiveness of the comprehensive decision result can be improved.

It should be further noted that, in order to improve accuracy of linear weighting, the above-mentioned using a linear weighting model to make a comprehensive decision on each weight value in the decision matrix, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and the corresponding comprehensive decision limiting threshold value, where the method includes: and comprehensively deciding each weight value by utilizing a linear weighting model, and determining a comprehensive decision value, wherein the linear weighting model is as follows: y _i is the integrated decision value of the ith sample, x _ij is the value of the jth feature quantity of the ith sample, and w _j is the weight value of the jth feature. The embodiment gives a specific linear weighting model, and improves the accuracy of linear weighting.

The fan blade fault detection method provided by the embodiment of the application can comprise the following steps: s101, acquiring the vibration acceleration of the blade, and performing fast Fourier transformation by using the vibration acceleration of the blade to determine the actual vibration mode frequency of the blade; s102, collecting axial vibration acceleration of a main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out fast Fourier transformation to determine axial actual rotation frequency of the main shaft and axial actual rotation frequency amplitude of the main shaft; s103, determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof. Therefore, according to the embodiment of the application, the actual blade vibration mode frequency, the main shaft axial actual rotation frequency and the main shaft axial actual rotation frequency amplitude are calculated by acquiring the data directly related to the fan blade, so that whether the fan blade has a fault or not is determined according to the actual blade vibration mode frequency, the main shaft axial actual rotation frequency and the main shaft axial actual rotation frequency amplitude. In addition, the embodiment can determine whether the blade mode frequency deviation exists or not according to the blade mode frequency deviation and the mode frequency deviation threshold value, and the blade mode frequency deviation can be used as one index for determining whether the fan blade has faults or not; in addition, the embodiment provides a specific method for determining the spindle axial actual rotation frequency and the spindle axial actual rotation frequency amplitude, so that the accuracy of the spindle axial actual rotation frequency and the amplitude determination thereof is improved; in addition, the embodiment provides a specific method for specifically calculating the axial actual rotation frequency of the main shaft and determining the unbalance of the wind wheel, so that the accuracy of determining the unbalance of the wind wheel is improved; in addition, the embodiment improves the accuracy of determining the fan blade faults by giving out specific steps of fan blade fault determination under the condition of blade icing; in addition, the embodiment gives specific parameters for judging mutation, and improves the accuracy of mutation determination of a trend curve of the frequency conversion-leaf frequency amplitude ratio; in addition, the embodiment can realize comprehensive decision on the fault state of the blade by utilizing an entropy method and linear weighting, so that the accuracy of determining the fault of the fan blade is improved; in addition, the number of the comprehensive decision limiting threshold values in the embodiment is multiple, the comprehensiveness of the comprehensive decision result can be improved, a specific linear weighting model is provided in the embodiment, and the accuracy of linear weighting is improved.

In order to facilitate understanding of the present invention, referring to fig. 2 specifically, fig. 2 is a flowchart illustrating a method for extracting a blade mode frequency feature according to an embodiment of the present invention, which may specifically include:

s201, vibration acceleration of the blade in a preset time is obtained by using a vibration sensor on the fan blade.

S202, integrating the vibration acceleration to obtain the vibration displacement of the blade.

This embodiment will perform a re-integration operation on the acceleration. The vibration sensor is additionally arranged on the wind turbine blade to acquire the vibration acceleration a (t) of the blade within a period of time t, and the displacement characteristic can reflect the vibration state of the blade because the wind turbine belongs to low-frequency rotating equipment, so that the vibration characteristic operation is carried out on the acceleration a (t) integrated into the displacement s (t). The expression for the integration of the acceleration a (t) into a single displacement sample s (t) is:

s(t)＝∫∫a(t)

s203, performing fast Fourier transform operation on the blade vibration displacement by using the blade mode frequency reference value to obtain the actual blade vibration mode frequency.

In the embodiment, the blade mode frequency reference value f _G (the range is 0-2 Hz) performs FFT operation on s (t), so as to avoid error in mode frequency searching caused by sampling precision, consider that actual mode frequency is searched within a certain range around theoretical mode frequency, and the processing procedure is as follows: setting Δf as an actual mode frequency searching frequency band f _b＝[f_G-Δf,f_G +Δf from left to right by using the center of f _G, and searching the maximum spectrum amplitude sf (g) of the displacement spectrum s (f _b) to obtain the actual blade vibration mode frequency f (g).

S204, taking the ratio of the difference between the actual blade vibration mode frequency and the theoretical blade vibration mode frequency as the blade mode frequency deviation.

Calculating blade modal frequency deviation

S205, judging whether the blade mode frequency deviation exceeds a frequency deviation threshold value, and determining whether the blade vibration mode frequency is changed.

And setting a limiting threshold value C ₁, judging whether the blade mode frequency deviation exceeds the limit, and when w ₁＞C₁, meeting the blade mode frequency deviation condition, wherein the range of the threshold value C1 is 0-0.1.

S206, when the frequency deviation threshold value is exceeded, the change is determined.

S207, when the frequency deviation threshold value is not exceeded, determining that no change occurs.

In order to facilitate understanding of the present invention, referring to fig. 3 in detail, fig. 3 is a flowchart illustrating a method for extracting unbalanced vibration characteristics of a wind turbine according to an embodiment of the present invention, which may specifically include:

S301, collecting axial vibration acceleration of the main shaft within preset time by using a vibration sensor below the main shaft of the wind turbine.

And (3) acquiring axial vibration acceleration of the main shaft by additionally arranging a vibration sensor below the main shaft of the wind turbine, and acquiring the axial vibration acceleration a _x (t) of the main shaft within a period of time t. As described above, the acceleration a (t) is integrated into the displacement s (t) and the vibration characteristic operation is performed to obtain the spindle axial vibration displacement s _x (t).

S302, integrating the axial vibration acceleration of the spindle to obtain axial vibration displacement of the spindle.

S303, acquiring wind wheel rotating speed data in the SCADA system in the same time period, calculating the average rotating speed of the wind wheel rotating speed in the time period, and determining the corresponding average rotating frequency of the wind wheel according to the average rotating speed of the wind wheel.

Acquiring wind wheel rotating speed data v (t) of a SCADA system in the same period t, and calculating the average wind wheel rotating speed of v (t) of the wind wheel rotating speed in the period tCorresponding wind wheel average rotation frequency f _n0,/>

More preferably, because the rotating speed of the wind wheel is low, the influence of the rotating speed fluctuation on the rotating frequency and the passing frequency of the blades is large, and if the real-time rotating speed signal of the wind wheel can be obtained, a more accurate rotating frequency value f _n0 in the monitoring period can be calculated.

S304, integral operation is carried out on the axial vibration displacement of the spindle, the maximum frequency spectrum amplitude is searched according to the average frequency spectrum of the wind wheel, and the spindle axial frequency and the amplitude thereof corresponding to the maximum frequency spectrum amplitude are determined as the spindle axial actual frequency and the spindle axial actual frequency amplitude.

And (3) carrying out FFT operation on s _x (t), and calculating to obtain the actual rotation frequency of the axial direction of the main shaft as f _nx and the actual rotation frequency amplitude of the axial direction of the main shaft as s _x(f_nx according to the method for obtaining the actual blade modal frequency.

S305, determining the blade vibration reference frequency by using the product of the number of fan blades and the actual rotation frequency vibration displacement of the axial direction of the main shaft.

If the number of the wind wheel blades is k, the reference frequency of the blade vibration is f _yp0＝k*f_nx, and in order to avoid error in blade frequency searching caused by sampling precision, the maximum frequency spectrum amplitude s _x(f_yp of the axial vibration displacement frequency spectrum of the main shaft is obtained according to the actual frequency searching method.

S306, carrying out actual frequency search on the blade vibration reference frequency by utilizing the average rotation frequency of the wind wheel to obtain the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft.

S307, determining the frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the actual frequency conversion amplitude of the axial direction of the main shaft and the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft.

S308, judging whether the amplitude ratio of the rotating frequency to the blade frequency exceeds a vibration threshold value, and determining whether unbalanced vibration of the wind wheel exists.

Calculating the ratio of the frequency conversion to the leaf frequency amplitude in the axial vibration displacement frequency spectrumAnd judging whether unbalanced vibration of the wind wheel exists or not by setting a related limiting threshold value C ₂, and when w ₂＞C₂, meeting the unbalanced vibration condition of the wind wheel.

And if the modal frequency of the blade and the unbalanced vibration characteristic value of the wind wheel exceed the limiting threshold, judging that the blade fault exists.

S309, when the wind wheel unbalance vibration exists, determining that the wind wheel unbalance vibration exists.

And S310, when the wind wheel unbalance vibration does not exceed the preset value, determining that the wind wheel unbalance vibration does not exist.

In order to facilitate understanding of the present invention, referring to fig. 4 specifically, fig. 4 is a flowchart illustrating a method for identifying a blade icing fault according to an embodiment of the present invention, which may specifically include:

S401, acquiring meteorological factors in a blade working environment; the meteorological factors include air temperature, air humidity and air pressure.

S402, determining whether the air temperature is smaller than or equal to a preset temperature threshold, whether the air humidity is larger than or equal to a preset humidity threshold and whether the air pressure is smaller than or equal to a preset air pressure threshold according to the preset temperature threshold, the preset humidity threshold, the preset air pressure threshold, the air temperature, the air humidity and the air pressure.

S403, when the air temperature is less than or equal to the preset temperature threshold value, the air humidity is greater than or equal to the preset humidity threshold value and the air pressure is less than or equal to the preset air pressure threshold value, the condition meeting the blade icing meteorological conditions is determined.

S404, if the modal frequency of the blade and the unbalance of the wind wheel are both overrun only under the blade icing meteorological condition, and the trend curve of the frequency conversion-blade frequency amplitude ratio has abrupt change, judging that the blade icing fault exists, otherwise, the blade icing fault does not belong to the non-icing fault.

This embodiment obtains meteorological factors including, but not limited to, air temperature T, air humidity H, and air pressure P _r in the blade operating environment and sets corresponding temperature threshold T _w, humidity threshold H _w, and air pressure threshold P _w.

The following 3 conditions are satisfied: ① The air temperature T is smaller than or equal to a preset temperature threshold T _w;②, and the air humidity H is larger than or equal to a preset humidity threshold Hw; ③ The air pressure P _r is smaller than or equal to a preset air pressure threshold P _w, and meets the blade icing meteorological conditions. The calculation formula of the trend mutation factor m about the frequency-to-leaf frequency amplitude ratio w ₂ is as follows: And setting a limiting threshold value C ₃, judging whether the unbalanced characteristic quantity is suddenly changed, and when m is more than C ₃, meeting a mutation condition.

If the modal frequency of the blade and the unbalance of the wind wheel are both out of limits only under the ice-coating meteorological condition of the blade, and the trend curve of the frequency-to-blade frequency amplitude ratio w ₂ has abrupt change, judging that the blade unbalance fault exists, otherwise, the blade unbalance fault belongs to the self fault of the blade and the non-ice-coating fault.

In the trend curve of the frequency conversion-leaf frequency amplitude ratio in this embodiment, there is a mutation in the ratio of the difference of the frequency conversion-leaf frequency amplitude ratio in the axial vibration displacement spectrum corresponding to each preset time point to the difference of each preset time point.

In order to facilitate understanding of the present invention, referring to fig. 5 in detail, fig. 5 is a flowchart illustrating a multi-feature comprehensive decision method according to an embodiment of the present invention, which may specifically include:

S501, forming a first feature matrix by feature value results of a plurality of continuous samples; wherein the value in the first feature matrix represents the value of the nth feature quantity of the mth sample.

In order to further realize accurate diagnosis, a first feature matrix D ₀ is formed by feature value results of a plurality of continuous samples, a second feature matrix D is obtained after the first feature matrix D ₀ is subjected to super-threshold calculation, the weight of each feature quantity in the second feature matrix D is calculated by adopting an entropy method, and comprehensive evaluation and decision on the fault state of the blade are realized based on linear weighting and threshold judgment. The method comprises the following specific steps:

And (3) comprehensive decision matrix calculation:

the expression of the first feature matrix D formed by calculating a plurality of feature values by a plurality of continuous samples is as follows:

Wherein X _ij represents the value of the j (j.ltoreq.n) th feature quantity of the ith (i.ltoreq.m) th sample, the implementation is not limited to a specific feature quantity type. For example, X ₁₁ in this embodiment may represent the value of the 1 st feature quantity of the 1 st sample, where the 1 st feature quantity is the blade mode frequency deviation, and X ₁₂ in this embodiment represents the value of the 2 nd feature quantity of the 1 st sample, where the 2 nd feature quantity is the conversion-to-blade frequency amplitude ratio in the axial vibration displacement spectrum; when it is required to determine whether the fan blade is faulty due to blade icing, the fan blade may also have a value that has X ₁₃,X₁₃ to indicate the 3 rd feature quantity of the 1 st sample, where the 3 rd feature quantity is a trend mutation factor.

S502, performing super-threshold calculation on the first feature matrix to obtain a second feature matrix.

The expression of the feature threshold C corresponding to the first feature matrix D is:

calculating the overrun of the first feature matrix D based on the feature threshold C to obtain a second feature matrix D, wherein the expression is as follows:

s503, calculating the weight of each feature quantity in the second feature matrix by adopting an entropy method.

S503 describes a weight value calculation process for calculating each element in the feature matrix, and the resulting weight is also a1×n matrix. The eigenvalue weights are determined based on entropy methods. Normalization processing is performed on the value of the j-th feature quantity of the i-th sample of the second feature matrix D:

Calculating entropy value of the j-th feature quantity:

Calculating the deviation degree of the j-th feature quantity:

d_j＝1-e_j

Normalizing the deviation index to obtain a weight value corresponding to the decision matrix:

/>

S504, determining a comprehensive parameter value based on linear weighting according to the weight of each characteristic quantity, and comparing the comprehensive parameter value with a plurality of comprehensive decision threshold values to realize comprehensive decision on the fault state of the blade.

And realizing comprehensive evaluation of the characteristic quantity based on a linear weighting method. After the weight of the characteristic factors is determined, comprehensively evaluating the characteristic factors by adopting a linear weighting model, wherein the evaluation model is as follows:

In the formula, y _i is the comprehensive grading value of the ith feature quantity, the full scale is 1, and the higher the grading is, the higher the credibility is. x _ij is the value of the j-th feature of the i-th sample, and w _j is the weight of the j-th feature.

And setting a plurality of comprehensive decision limiting thresholds such as early warning, alarming and the like according to actual demands, and outputting a comprehensive decision result by judging the magnitude of the comprehensive score value y _i and the threshold of the sample.

Aiming at the blade icing fault example, a second feature matrix D formed by 3 feature variable values obtained by calculating 3 continuous samples is set as follows:

According to the above S503 calculation process, the matrix obtained after normalization processing is

Each characteristic entropy value matrix is obtained as follows:

w＝[0.99 0.99 0.99]

The entropy value deviation matrix is obtained as follows:

d＝[0.01 0.01 0.01]

the weight matrix of the 3 feature variables is further obtained as follows:

w＝[3/10 3/10 3/10]

And finally, calculating the comprehensive grading value of each characteristic quantity by adopting linear weighting.

y＝[0.186 0.174 0.159]

And comparing the comprehensive scoring values of the characteristic quantities in the comprehensive scoring matrix with a preset pre-warning threshold value, and outputting the corresponding blade icing fault degree through comprehensive decision.

The embodiment provides a method for realizing the on-line identification and alarm decision-making of the wind turbine blade faults based on the multi-vibration characteristic joint diagnosis, which comprises the steps of firstly identifying the blade faults by extracting a plurality of vibration characteristics of the blade, then adopting a multi-sample and multi-characteristic entropy weighting comprehensive decision-making method, further realizing quantitative analysis and accurate diagnosis of the fault, avoiding the false decision caused by an abnormal single sample and integrally realizing the on-line diagnosis of the wind turbine blade faults. Through failure mechanism analysis, the online monitoring and accurate diagnosis of the failure of the wind turbine blade are realized without depending on a large amount of historical data.

The following describes a fan blade fault detection device provided in the embodiment of the present invention, and the fan blade fault detection device described below and the fan blade fault detection method described above may be referred to correspondingly.

Referring to fig. 6 specifically, fig. 6 is a schematic structural diagram of a fan blade fault detection device according to an embodiment of the present invention, which may include:

the actual blade vibration mode frequency determining module 100 is configured to collect a blade vibration acceleration, and perform a fast fourier transform to determine an actual blade vibration mode frequency using the blade vibration acceleration;

The main shaft axial actual rotation frequency determining module 200 is configured to collect main shaft axial vibration acceleration, and perform the fast fourier transform to determine main shaft axial actual rotation frequency and main shaft axial actual rotation frequency amplitude by using the main shaft axial vibration acceleration;

the blade fault determining module 300 is configured to determine whether the fan blade has a fault by using the actual blade vibration mode frequency, the spindle axial actual rotation frequency, and the amplitude thereof.

Based on the above embodiment, the actual blade vibration mode frequency determining module 100 may include:

the blade vibration displacement determining unit is used for carrying out integral operation on the blade vibration acceleration to determine the blade vibration displacement;

And the actual blade vibration mode frequency determining unit is used for determining a first preset searching range by utilizing the theoretical blade vibration mode frequency after carrying out the fast Fourier transform operation on the blade vibration displacement, searching the maximum frequency spectrum amplitude in the first preset searching range, and determining the blade vibration mode frequency corresponding to the maximum frequency spectrum amplitude as the actual blade vibration mode frequency.

Further, based on any of the above embodiments, the above blade failure determination module 300 may include:

The blade mode frequency deviation calculation unit is used for dividing the theoretical blade mode frequency by the difference value of the actual blade vibration mode frequency and the theoretical blade vibration mode frequency to obtain blade mode frequency deviation;

and the blade modal frequency deviation determining unit is used for determining whether the blade modal frequency deviation exists or not by utilizing the blade modal frequency deviation and the modal frequency deviation threshold value.

Further, based on any of the above embodiments, the above spindle axial actual rotation frequency determining module 200 may include:

the main shaft axial vibration displacement determining unit is used for carrying out integral operation on the main shaft axial vibration acceleration to obtain main shaft axial vibration displacement;

the main shaft axial actual rotation frequency and main shaft axial actual rotation frequency amplitude determining unit is used for carrying out the fast Fourier transform operation on the main shaft axial vibration displacement, determining a second preset searching range by utilizing wind wheel average rotation frequency, searching the maximum frequency spectrum amplitude in the second preset searching range, and determining the main shaft axial rotation frequency and the amplitude thereof corresponding to the maximum frequency spectrum amplitude as the main shaft axial actual rotation frequency and the main shaft axial actual rotation frequency amplitude.

Further, based on any of the above embodiments, the fan blade failure detection apparatus may further include:

a main shaft axial vibration displacement frequency spectrum maximum frequency spectrum amplitude determining module for determining a blade vibration reference frequency by utilizing the product of the main shaft axial actual rotation frequency and the number of blades,

Determining a third preset searching range by utilizing the blade vibration reference frequency, searching the maximum frequency spectrum amplitude in the third preset searching range, and determining the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft;

The frequency conversion-leaf frequency amplitude ratio determining module is used for determining the frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the actual frequency conversion amplitude of the axial direction of the main shaft to the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft;

and the wind wheel unbalanced vibration condition determining module is used for determining whether the wind wheel unbalanced vibration condition is met according to the frequency conversion-blade frequency amplitude ratio and the frequency conversion-blade frequency amplitude threshold value.

Further, based on the above embodiment, the fan blade failure detection apparatus may further include:

The three condition determining modules are used for determining whether the blade modal frequency deviation exceeds a threshold, whether the frequency-to-blade frequency amplitude exceeds the threshold and whether abrupt changes exist in a trend curve of the frequency-to-blade frequency amplitude ratio under the blade icing meteorological conditions when the blade icing meteorological conditions are met;

And the blade icing fault determination module is used for determining that the blade icing fault exists if the blade modal frequency deviation exceeds a threshold, the frequency conversion-blade frequency amplitude ratio exceeds a threshold and a mutation exists in a trend curve of the frequency conversion-blade frequency amplitude ratio.

Further, based on the above embodiment, the fan blade failure detection apparatus may further include:

The meteorological factor acquisition module is used for acquiring meteorological factors in the working environment of the blade; wherein the meteorological factors include air temperature, air humidity and air pressure;

The weather judging module is used for determining whether the air temperature is smaller than or equal to the preset temperature threshold value, whether the air humidity is larger than or equal to the preset humidity threshold value and whether the air pressure is smaller than or equal to the preset air pressure threshold value according to the preset temperature threshold value, the preset humidity threshold value and the preset air pressure threshold value;

And the blade icing meteorological conditions determining module is used for determining that the blade icing meteorological conditions are met when the air temperature is smaller than or equal to the preset temperature threshold, the air humidity is larger than or equal to the preset humidity threshold and the air pressure is smaller than or equal to the preset air pressure threshold.

Further, the determining process of the mutation in the trend curve of the frequency conversion-leaf frequency amplitude ratio may include:

The trend mutation factor determining unit is used for determining trend mutation factors corresponding to the two time points; wherein the formula of the trend mutation factor is Wherein i and k respectively represent different time points, w ₂ (i) represents the frequency conversion-leaf frequency amplitude ratio of the i point, and w ₂ (k) represents the frequency conversion-leaf frequency amplitude ratio of the k point;

the judgment unit is used for determining whether the trend mutation factor is larger than a trend mutation threshold value or not based on the trend mutation threshold value;

and a mutation determining unit for determining that a mutation exists when the mutation is larger than the threshold value.

Further, based on any of the above embodiments, the above blade failure determination module 300 may include:

A first type characteristic quantity determining unit, configured to determine a blade mode frequency deviation by using a ratio of a difference between the actual blade vibration mode frequency and the theoretical blade mode frequency to the theoretical blade mode frequency, and use the blade mode frequency deviation as a first type characteristic quantity;

The second type characteristic quantity determining unit is used for determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the spindle axial actual frequency conversion amplitude of the spindle axial actual frequency conversion to the maximum frequency spectrum amplitude of the spindle axial vibration displacement frequency spectrum, and taking the frequency conversion-leaf frequency amplitude ratio as a second type characteristic quantity;

a first feature matrix constructing unit configured to construct a first feature matrix using a plurality of the first type feature amounts and a plurality of the second type feature amounts corresponding to a plurality of consecutive samples; wherein the elements in the first feature matrix represent the values of the nth class feature quantity of the mth sample;

the second feature matrix determining unit is used for obtaining a second feature matrix after performing super-threshold calculation on the first feature matrix;

And the comprehensive decision unit is used for calculating the weight of each characteristic quantity in the second characteristic matrix by adopting an entropy method and realizing comprehensive decision on the fault state of the blade based on a linear weighting model and each characteristic threshold value.

Further, based on the above embodiment, the above integrated decision unit may include:

the normalization processing unit is used for performing normalization processing on the values of all the characteristic quantities of each sample of the second characteristic matrix;

The deviation degree determining unit is used for calculating entropy values of all the characteristic quantities by utilizing an entropy value method and determining the deviation degree by utilizing the entropy values;

a weight value determining unit for normalizing the deviation degree to obtain weight values of various characteristic quantities,

The decision matrix construction unit is used for constructing the weight values to obtain a decision matrix;

and the comprehensive decision unit is used for comprehensively deciding each weight value in the decision matrix by utilizing the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and the corresponding comprehensive decision limiting threshold value.

Further, based on the above embodiment, the above integrated decision unit may include:

the linear weighting model based comprehensive decision unit is used for comprehensively deciding each weight value by utilizing the linear weighting model to determine a comprehensive decision value, and the linear weighting model is as follows: y _i is the integrated decision value of the ith sample, x _ij is the value of the jth feature quantity of the ith sample, and w _j is the weight value of the jth feature.

Further, based on any of the above embodiments, the blade vibration acceleration is obtained by a sensor on the fan blade; and collecting axial vibration acceleration of the main shaft through a vibration sensor below the main shaft of the fan.

It should be noted that, the modules and units in the fan blade failure detection device can change the sequence of the modules and units before and after the modules and units do not affect the logic.

The fan blade fault detection device provided by the embodiment of the application can comprise: the actual blade vibration mode frequency determining module 100 is configured to collect a blade vibration acceleration, and perform a fast fourier transform to determine an actual blade vibration mode frequency using the blade vibration acceleration; the main shaft axial actual rotation frequency determining module 200 is configured to collect main shaft axial vibration acceleration, and perform the fast fourier transform to determine main shaft axial actual rotation frequency and main shaft axial actual rotation frequency amplitude by using the main shaft axial vibration acceleration; the blade fault determining module 300 is configured to determine whether the fan blade has a fault by using the actual blade vibration mode frequency, the spindle axial actual rotation frequency, and the amplitude thereof. Therefore, according to the embodiment of the application, the actual blade vibration mode frequency, the main shaft axial actual rotation frequency and the main shaft axial actual rotation frequency amplitude are calculated by acquiring the data directly related to the fan blade, so that whether the fan blade has a fault or not is determined according to the actual blade vibration mode frequency, the main shaft axial actual rotation frequency and the main shaft axial actual rotation frequency amplitude. In addition, the embodiment can determine whether the blade mode frequency deviation exists or not according to the blade mode frequency deviation and the mode frequency deviation threshold value, and the blade mode frequency deviation can be used as one index for determining whether the fan blade has faults or not; in addition, the embodiment provides a method for specifically determining the spindle axial actual rotation frequency and the spindle axial actual rotation frequency amplitude, so that the accuracy of the spindle axial actual rotation frequency and the amplitude determination thereof is improved; in addition, the embodiment provides a specific method for specifically calculating the axial actual rotation frequency of the main shaft and determining the unbalance of the wind wheel, so that the accuracy of determining the unbalance of the wind wheel is improved; in addition, the embodiment improves the accuracy of determining the fan blade faults by giving out specific steps of fan blade fault determination under the condition of blade icing; in addition, the embodiment gives specific parameters for judging mutation, and improves the accuracy of mutation determination of a trend curve of the frequency conversion-leaf frequency amplitude ratio; in addition, the embodiment can realize comprehensive decision on the fault state of the blade by utilizing an entropy method and linear weighting, so that the accuracy of determining the fault of the fan blade is improved; in addition, the number of the comprehensive decision limiting threshold values in the embodiment is multiple, the comprehensiveness of the comprehensive decision result can be improved, a specific linear weighting model is provided in the embodiment, and the accuracy of linear weighting is improved.

The following describes a fan blade fault detection device provided in an embodiment of the present invention, where the fan blade fault detection device described below and the fan blade fault detection method described above may be referred to correspondingly.

Referring to fig. 7, fig. 7 is a schematic structural diagram of a fan blade fault detection device according to an embodiment of the present invention, which may include:

A memory 10 for storing a computer program;

And a processor 20 for executing a computer program to implement the fan blade failure detection method described above.

The memory 10, the processor 20, and the communication interface 30 all communicate with each other via a communication bus 40.

In the embodiment of the present invention, the memory 10 is used for storing one or more programs, the programs may include program codes, the program codes include computer operation instructions, and in the embodiment of the present invention, the memory 10 may store programs for implementing the following functions:

Acquiring the vibration acceleration of the blade, and determining the actual vibration mode frequency of the blade by utilizing the vibration acceleration of the blade to perform fast Fourier transformation;

Collecting axial vibration acceleration of the main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out fast Fourier transformation to determine axial actual rotation frequency of the main shaft and axial actual rotation frequency amplitude of the main shaft;

And determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the actual spindle axial rotation frequency and the amplitude thereof.

In one possible implementation, the memory 10 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, and at least one application program required for functions, etc.; the storage data area may store data created during use.

In addition, memory 10 may include read only memory and random access memory and provide instructions and data to the processor. A portion of the memory may also include NVRAM. The memory stores an operating system and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, where the operating instructions may include various operating instructions for performing various operations. The operating system may include various system programs for implementing various basic tasks as well as handling hardware-based tasks.

Processor 20 may be a Central processing unit (Central ProcessingUnit, CPU), an asic, a dsp, a fpga or other programmable logic device, and processor 20 may be a microprocessor or any conventional processor. The processor 20 may call a program stored in the memory 10.

The communication interface 30 may be an interface of a communication module for connecting with other devices or systems.

Of course, it should be noted that the structure shown in fig. 7 is not limited to the fan blade failure detection apparatus according to the embodiment of the present invention, and the fan blade failure detection apparatus may include more or fewer components than those shown in fig. 7, or may combine some components in practical applications.

The following describes a computer readable storage medium provided in an embodiment of the present invention, where the computer readable storage medium described below and the fan blade fault detection method described above may be referred to correspondingly.

The invention also provides a readable storage medium, wherein the readable storage medium stores a computer program, and the computer program realizes the steps of the fan blade fault detection method when being executed by a processor.

The readable storage medium may include: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, so that the same or similar parts between the embodiments are referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.

Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

Finally, it is further noted that, in this document, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.

The method, the device, the equipment and the readable storage medium for detecting the faults of the fan blades provided by the invention are described in detail, and specific examples are applied to the explanation of the principle and the implementation mode of the invention, and the explanation of the examples is only used for helping to understand the method and the core idea of the invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.

Claims (15)

1. A fan blade failure detection method, comprising:

collecting the vibration acceleration of the blade, and performing fast Fourier transformation by using the vibration acceleration of the blade to determine the actual vibration mode frequency of the blade;

collecting axial vibration acceleration of a main shaft, and utilizing the axial vibration acceleration of the main shaft to carry out the fast Fourier transformation to determine the axial actual rotation frequency of the main shaft and the axial actual rotation frequency amplitude of the main shaft;

And determining whether the fan blade has faults or not by using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof.

2. The fan blade failure detection method of claim 1, wherein the collecting blade vibration acceleration and performing a fast fourier transform using the blade vibration acceleration to determine an actual blade vibration mode frequency comprises:

Performing integral operation on the blade vibration acceleration to determine blade vibration displacement;

After the fast fourier transform operation is performed on the blade vibration displacement, determining a first preset search range by utilizing theoretical blade vibration modal frequency, searching the maximum spectrum amplitude in the first preset search range, and determining the blade vibration modal frequency corresponding to the maximum spectrum amplitude as the actual blade vibration modal frequency.

3. The fan blade failure detection method according to claim 1, wherein the determining whether the fan blade has a failure using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof includes:

Dividing the difference value of the actual blade vibration mode frequency and the theoretical blade vibration mode frequency by the theoretical blade mode frequency to obtain blade mode frequency deviation;

and determining whether the blade modal frequency deviation exists or not by using the blade modal frequency deviation and the modal frequency deviation threshold value.

4. The fan blade failure detection method according to claim 1, wherein the collecting the spindle axial vibration acceleration and performing the fast fourier transform using the spindle axial vibration acceleration to determine a spindle axial actual rotation frequency and a spindle axial actual rotation frequency amplitude includes:

Performing integral operation on the axial vibration acceleration of the spindle to obtain axial vibration displacement of the spindle;

And carrying out the fast Fourier transform operation on the axial vibration displacement of the spindle, determining a second preset searching range by utilizing the average frequency conversion of the wind wheel, searching the maximum frequency spectrum amplitude in the second preset searching range, and determining the spindle axial frequency conversion corresponding to the maximum frequency spectrum amplitude and the amplitude thereof as the spindle axial actual frequency conversion and the spindle axial actual frequency conversion amplitude.

5. The fan blade failure detection method according to claim 1, wherein the determining whether the fan blade has a failure using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof includes:

Determining a blade vibration reference frequency by using the product of the actual axial rotation frequency of the spindle and the number of blades, determining a third preset search range by using the blade vibration reference frequency, searching the maximum frequency spectrum amplitude in the third preset search range, and determining the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the spindle;

Determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the actual frequency conversion amplitude of the axial direction of the main shaft to the maximum frequency spectrum amplitude of the axial vibration displacement frequency spectrum of the main shaft;

And determining whether the unbalanced vibration condition of the wind wheel is met according to the frequency conversion-blade frequency amplitude ratio and the frequency conversion-blade frequency amplitude threshold value.

6. The fan blade failure detection method of claim 5, further comprising:

Determining whether the blade modal frequency deviation exceeds a threshold, whether the frequency-to-blade frequency amplitude exceeds the threshold and whether abrupt changes exist in a trend curve of the frequency-to-blade frequency amplitude ratio under the blade icing meteorological conditions when the blade icing meteorological conditions are met;

And if the blade mode frequency deviation exceeds a threshold, the frequency conversion-blade frequency amplitude ratio exceeds the threshold, and a mutation exists in a trend curve of the frequency conversion-blade frequency amplitude ratio, determining that the blade icing fault exists.

7. The method of claim 6, further comprising, prior to said determining that blade icing weather conditions are met:

Acquiring meteorological factors in a blade working environment; wherein the meteorological factors include air temperature, air humidity and air pressure;

Determining whether the air temperature is less than or equal to the preset temperature threshold, whether the air humidity is greater than or equal to the preset humidity threshold, and whether the air pressure is less than or equal to the preset air pressure threshold according to a preset temperature threshold, a preset humidity threshold and a preset air pressure threshold;

And when the air temperature is determined to be less than or equal to the preset temperature threshold value, the air humidity is determined to be greater than or equal to the preset humidity threshold value, and the air pressure is determined to be less than or equal to the preset air pressure threshold value, determining that the blade icing meteorological conditions are met.

8. The fan blade failure detection method according to claim 5, wherein the determining that there is a mutation in the trend curve of the rotor-to-blade-to-frequency-to-amplitude ratio includes:

determining trend mutation factors corresponding to the two time points; wherein the formula of the trend mutation factor is Wherein i and k respectively represent different time points, w ₂ (i) represents the frequency conversion-leaf frequency amplitude ratio of the i point, and w ₂ (k) represents the frequency conversion-leaf frequency amplitude ratio of the k point;

determining whether the trend mutation factor is greater than a trend mutation threshold value;

when greater than, the presence of a mutation is determined.

9. The fan blade failure detection method according to any one of claims 1 to 8, wherein the determining whether the fan blade has a failure using the actual blade vibration mode frequency and the spindle axial actual rotation frequency and the amplitude thereof includes:

Determining a blade mode frequency deviation by utilizing the ratio of the difference between the actual blade vibration mode frequency and the theoretical blade mode frequency to the theoretical blade mode frequency, and taking the blade mode frequency deviation as a first type of characteristic quantity;

Determining a frequency conversion-leaf frequency amplitude ratio by utilizing the ratio of the spindle axial actual frequency conversion amplitude of the spindle axial actual frequency conversion to the maximum frequency spectrum amplitude of the spindle axial vibration displacement frequency spectrum, and taking the frequency conversion-leaf frequency amplitude ratio as a second type of characteristic quantity;

constructing a first feature matrix by utilizing a plurality of first type feature quantities and a plurality of second type feature quantities corresponding to a plurality of continuous samples; wherein the elements in the first feature matrix represent the values of the nth class feature quantity of the mth sample;

performing super-threshold calculation on the first feature matrix to obtain a second feature matrix;

And calculating the weight of each characteristic quantity in the second characteristic matrix by adopting an entropy method, and realizing comprehensive decision on the fault state of the blade based on a linear weighting model and each characteristic threshold value.

10. The fan blade failure detection method according to claim 9, wherein the calculating the weight of each feature in the second feature matrix by using an entropy method, and implementing the comprehensive decision on the blade failure state based on the linear weighting model and each feature threshold value includes:

Normalizing the values of all the feature quantities of each sample of the second feature matrix;

calculating entropy values of all the characteristic quantities by using an entropy method, and determining deviation degree by using the entropy values;

normalizing the deviation degree to obtain weight values of various characteristic quantities,

Constructing the weight value to obtain a decision matrix;

And carrying out comprehensive decision on each weight value in the decision matrix by utilizing the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and the corresponding comprehensive decision limiting threshold value.

11. The fan blade fault detection method according to claim 10, wherein the performing a comprehensive decision on each weight value in the decision matrix by using the linear weighting model, determining a comprehensive decision value, and outputting a comprehensive decision result by judging the magnitude of the comprehensive decision value of each sample and a corresponding comprehensive decision limiting threshold value includes:

And carrying out comprehensive decision on each weight value by using the linear weighting model to determine a comprehensive decision value, wherein the linear weighting model is as follows: y _i is the integrated decision value of the ith sample, x _ij is the value of the jth feature quantity of the ith sample, and w _j is the weight value of the jth feature.

12. The fan blade failure detection method according to any one of claims 1 to 8, wherein the blade vibration acceleration is obtained by a sensor on a fan blade; and collecting axial vibration acceleration of the main shaft through a vibration sensor below the main shaft of the fan.

13. A fan blade failure detection apparatus, comprising:

The actual blade vibration mode frequency determining module is used for collecting the blade vibration acceleration and performing fast Fourier transformation by utilizing the blade vibration acceleration to determine the actual blade vibration mode frequency;

the main shaft axial actual rotation frequency determining module is used for collecting main shaft axial vibration acceleration and utilizing the main shaft axial vibration acceleration to carry out the fast Fourier transformation to determine main shaft axial actual rotation frequency and main shaft axial actual rotation frequency amplitude;

And the blade fault determining module is used for determining whether the fan blade has faults or not by utilizing the actual blade vibration modal frequency, the spindle axial actual rotation frequency and the amplitude thereof.

14. A fan blade failure detection apparatus, comprising:

A memory for storing a computer program;

A processor for implementing the fan blade failure detection method according to any of claims 1 to 12 when executing the computer program.

15. A readable storage medium having stored therein computer executable instructions which when loaded and executed by a processor implement the fan blade failure detection method of any of claims 1 to 12.