CN109973325B - Method and apparatus for identifying abnormal vibration - Google Patents

Abstract

A method and device for identifying abnormal vibrations are provided, the method comprising: obtaining operating data of a predetermined component of a wind turbine generator set in multiple time periods, the operating data comprising vibration acceleration data of the predetermined component and rotation speed data related to the predetermined component; performing frequency domain conversion on the vibration acceleration data of the predetermined component in the multiple time periods to obtain multiple corresponding acceleration spectra; determining the frequency value used for abnormal vibration analysis in each acceleration spectrum; determining whether the predetermined component has abnormal vibration based on the determined frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotation speed data related to the predetermined component in each time period. The method and device for identifying abnormal vibrations according to the exemplary embodiment of the present invention can timely and accurately locate the predetermined component in the wind turbine generator set that has abnormal vibrations, providing strong support for effectively evaluating the vibration state of the predetermined component.

Description

Method and apparatus for identifying abnormal vibration

technical field

The present invention relates to the technical field of wind power generation, and more particularly, to a method and device for identifying abnormal vibration.

Background technique

Wind turbines are generally installed in remote wind farms, and the reliability and stability of the main components in the wind turbine are very important. In order to enable the equipment to operate safely, stably, long-term, and at full load, it is necessary to keep abreast of the operating status of the equipment, prevent failures, eliminate accidents, extend the operating cycle of the equipment, shorten the maintenance time, maximize the production potential of the equipment, and prevent problems in time. The main components in the generator set are safe, stable, indefinite period, and the operating conditions of changing loads are controlled in a timely manner.

In the long-term operation, due to the special geographical location and operating environment of the wind farm, and the concealment of the problem, it is difficult for the on-site staff to discover the hidden safety hazards of the main components of the wind turbine during the power generation process (such as non-obvious abnormality) in time. vibration). Among them, the fundamental frequency of the wind turbine and its frequency-doubling vibration are common vibration forms, which will bring different degrees of damage to the wind turbine and even the entire wind turbine, and will lead to the failure of the wind turbine in severe cases.

The general algorithm of the existing fundamental wave frequency identification method is relatively complex, requires a lot of calculation and signal analysis, and is inefficient when used for the identification of large batches of data.

SUMMARY OF THE INVENTION

The purpose of the exemplary embodiments of the present invention is to provide a method and device for identifying abnormal vibration, so as to solve the technical problems in the prior art that the abnormal vibration of the main components in the wind turbine cannot be found in time, and the identification efficiency of the abnormal vibration is low. .

According to an aspect of an exemplary embodiment of the present invention, there is provided a method for identifying abnormal vibration, the method comprising: acquiring operation data of predetermined components of a wind turbine in a plurality of time periods, the operation data including the predetermined vibration acceleration data of the component and rotational speed data related to the predetermined component; frequency domain conversion is performed on the vibration acceleration data of the predetermined component in the multiple time periods, respectively, to obtain a plurality of corresponding acceleration spectra; determine The frequency value for abnormal vibration analysis in each acceleration spectrum; the predetermined component is determined based on the determined frequency value for abnormal vibration analysis in each acceleration spectrum and rotational speed data related to the predetermined component in each time period Is there any abnormal vibration.

Optionally, the type of abnormal vibration may include abnormal fundamental frequency vibration of the predetermined component and abnormal multiplied frequency vibration of the predetermined component.

Optionally, the step of determining the frequency value used for abnormal vibration analysis in each acceleration spectrum may include: searching for a frequency point in the acceleration spectrum whose frequency amplitude value is greater than a frequency amplitude threshold; The frequency value is used as the frequency value in the acceleration spectrum for abnormal vibration analysis.

Optionally, the step of determining the frequency value used for abnormal vibration analysis in each acceleration spectrum may include: determining whether the frequency amplitude value corresponding to the preset frequency of interest in the acceleration spectrum is greater than the frequency amplitude threshold; If the frequency amplitude value corresponding to the preset frequency of interest is greater than the frequency amplitude threshold, the frequency value corresponding to the preset frequency of interest is used as the frequency value in the acceleration spectrum for abnormal vibration analysis.

Optionally, the preset frequency point of interest may be a predetermined number of frequency points before all frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value.

Optionally, the step of determining whether the predetermined component has abnormal vibration based on the determined frequency value for abnormal vibration analysis in each acceleration spectrum and rotational speed data related to the predetermined component in each time period may include: : determine whether the predetermined linear distribution law is satisfied between the frequency value and the rotational speed data; when the predetermined linear distribution law is satisfied between the frequency value and the rotational speed data, determine that the predetermined component has abnormal vibration.

Optionally, the step of determining whether a predetermined linear distribution law is satisfied between the frequency value and the rotational speed data may include: drawing a rotational speed-frequency scatterplot based on the frequency value and the rotational speed data, wherein the A scatter point in the rotational speed-frequency scattergram can correspond to rotational speed data of a time period and a frequency value used for abnormal vibration analysis in the acceleration spectrum corresponding to the time period; select a preset range within a predetermined frequency linear model scatter points within the scatter point; determine whether the predetermined linear distribution law of the predetermined frequency linear model is satisfied between the frequency value and the rotational speed data based on the selected scatter points.

Optionally, the step of determining whether the relationship between the frequency value and the rotational speed data satisfies the predetermined linear distribution law of the predetermined frequency linear model based on the selected scatter points may include: Perform linear regression on the frequency value and the rotational speed data to obtain the model parameters of the linear regression; calculate the difference between the model parameter and the specified parameter of the predetermined component; when the difference is not greater than the first set value, It is determined that a predetermined linear distribution law of the predetermined frequency linear model is satisfied between the frequency value corresponding to the selected scatter point and the rotational speed data.

Optionally, the step of determining whether the predetermined linear distribution law of the predetermined frequency linear model is satisfied between the frequency value and the rotational speed data based on the selected scatter points may include: establishing an objective function, the objective function indicating The distance from each scatter point to the predetermined frequency linear model; the value of the objective function is obtained by substituting the frequency value corresponding to the selected scatter point and the rotational speed data into the objective function; when the value of the objective function is When the value is not greater than the set value, it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data satisfy the predetermined linear distribution law of the predetermined frequency linear model.

Optionally, the step of determining whether a predetermined linear distribution law is satisfied between the frequency value and the rotational speed data may include: calculating, based on the rotational speed data, a rotational speed statistical value reflecting data characteristics for each time period; Whether the predetermined linear distribution law is satisfied between the frequency value and the rotational speed statistical value.

Optionally, the rotational speed statistic value reflecting data characteristics in each time period may include any one of the following items: the average value of rotational speed data related to the predetermined component in the time period, the The median value of the rotational speed data related to the predetermined component, and the effective value of the rotational speed data related to the predetermined component within the time period.

Optionally, the step of performing frequency domain conversion on the vibration acceleration data of the predetermined component in the plurality of time periods to obtain a plurality of corresponding acceleration spectra may include: determining the relationship between each time period and the Whether the rotational speed data related to the predetermined component is within the set rotational speed range; if the rotational speed data related to the predetermined component is within the set rotational speed range within any period of time, then the predetermined component is within the specified rotational speed range within the specified period of time. The vibration acceleration data are converted in frequency domain to obtain a corresponding acceleration spectrum.

Optionally, the step of determining whether the rotational speed data related to the predetermined component in each time period is within a set rotational speed range may include: calculating a criterion for the rotational speed data related to the predetermined component in each time period difference or steady-state error; if the calculated standard deviation or steady-state error for any period of time is within a set threshold range, it is determined that the rotational speed data related to the predetermined component in said any period of time is within the set rotational speed range If the standard deviation or steady state error calculated for any time period is not within the set threshold range, it is determined that the rotational speed data related to the predetermined component in the any time period is not within the set rotational speed range.

Optionally, the predetermined component may include any one of the following items: a wind turbine, a generator cogging, a gear box, and a rolling bearing, wherein, when the predetermined component is a wind turbine, the predetermined component The vibration acceleration data may be the vibration acceleration data of the nacelle of the wind turbine, the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine, and the type of abnormal vibration may include the fundamental wave frequency vibration abnormality and the fundamental wave of the wind turbine. The frequency double frequency vibration is abnormal; when the predetermined component is a generator cogging, the vibration acceleration data of the predetermined component can be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component can be wind The rotational speed data of the generator, the types of abnormal vibration may include abnormal vibration of the cogging frequency of the generator and abnormal vibration of the frequency multiplier of the cogging frequency of the generator; when the predetermined component is a gearbox, the vibration acceleration data of the predetermined component may be is the vibration acceleration data of the power gear or driven gear in the gearbox, the rotational speed data related to the predetermined component may be the rotational speed data of the shaft where the power gear or the driven gear is located, and the type of abnormal vibration may include the meshing frequency of the power gear and Abnormal frequency doubled vibration of the meshing frequency of the power gear, or abnormal frequency doubled vibration of the meshing frequency of the driven gear and the meshing frequency of the driven gear; when the predetermined component is a rolling bearing, the vibration acceleration data of the predetermined component may be the bearing of the rolling bearing The vibration acceleration data of the seat, the rotational speed data related to the predetermined component may be the rotational speed data of the rolling bearing, and the types of abnormal vibration may include abnormal vibration of the characteristic frequency of the rolling bearing fault and abnormal vibration of the double frequency of the characteristic frequency of the rolling bearing fault.

According to another aspect of an exemplary embodiment of the present invention, there is provided a device for identifying abnormal vibration, the device comprising: an operation data acquisition module configured to acquire operation data of predetermined components of a wind turbine in a plurality of time periods, The operation data includes vibration acceleration data of the predetermined component and rotational speed data related to the predetermined component; the time-frequency conversion module is configured to perform the vibration acceleration data of the predetermined component in the multiple time periods respectively. Frequency domain conversion to obtain a plurality of acceleration spectra corresponding respectively; a frequency value determination module for determining the frequency value in each acceleration spectrum for abnormal vibration analysis; an abnormal vibration analysis module for each acceleration spectrum determined based on The frequency value used in the abnormal vibration analysis and the rotational speed data related to the predetermined component in each time period determine whether the predetermined component has abnormal vibration.

Optionally, the type of abnormal vibration may include abnormal fundamental frequency vibration of the predetermined component and abnormal multiplied frequency vibration of the predetermined component.

Optionally, the frequency value determination module can be used to find a frequency point in the acceleration spectrum where the frequency amplitude value is greater than the frequency amplitude threshold value, and use the frequency value corresponding to the found frequency point as the frequency value used for abnormal vibration analysis in the acceleration spectrum. frequency value.

Optionally, the frequency value determination module can be used to determine whether the frequency amplitude value corresponding to the preset frequency of interest in the acceleration spectrum is greater than the frequency amplitude threshold, if the frequency amplitude value corresponding to the preset frequency of interest is greater than the frequency. the amplitude threshold, the frequency value determination module takes the frequency value corresponding to the preset frequency point of interest as the frequency value in the acceleration spectrum for abnormal vibration analysis.

Optionally, the preset frequency point of interest may be a predetermined number of frequency points before all frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value.

Optionally, the abnormal vibration analysis module can be used to determine whether a predetermined linear distribution law is satisfied between the frequency value and the rotational speed data, and when the frequency value and the rotational speed data satisfy the predetermined linear distribution law. , it is determined that the predetermined component has abnormal vibration.

Optionally, the abnormal vibration analysis module may include: a scatterplot drawing module, configured to draw a rotational speed-frequency scattergram based on the frequency value and the rotational speed data, wherein the rotational speed-frequency scattergram is A scatter point can correspond to the rotational speed data of a period of time and a frequency value used for abnormal vibration analysis in the acceleration spectrum corresponding to the period of time; the scatter point screening module is used to select a frequency value within the preset range of the predetermined frequency linear model The scatter point; a linear distribution determination module, configured to determine, based on the selected scatter point, whether the relationship between the frequency value and the rotational speed data satisfies the predetermined linear distribution law of the predetermined frequency linear model.

Optionally, the linear distribution determination module may be configured to perform linear regression on the frequency value corresponding to the selected scatter point and the rotational speed data, obtain model parameters of the linear regression, and calculate the relationship between the model parameters and the predetermined value. The difference value of the specified parameter of the component, when the difference value is not greater than the first set value, it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data satisfy the predetermined frequency linear model Linear distribution law.

Optionally, the linear distribution determination module may include: an objective function establishment sub-module for establishing an objective function, the objective function indicating the root mean square value of the distance from each scatter point to the predetermined frequency linear model; A function value calculation sub-module for obtaining the value of the objective function by substituting the frequency value and the rotational speed data corresponding to the selected scatter points into the objective function; a distribution law determination sub-module for when the target When the value of the function is not greater than the set value, it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data satisfy the predetermined linear distribution law of the predetermined frequency linear model.

Optionally, the abnormal vibration analysis module can be used to calculate a rotational speed statistical value reflecting data characteristics of each time period based on the rotational speed data, and determine whether the frequency value and the rotational speed statistical value satisfy a predetermined linear distribution. law.

Optionally, the rotational speed statistic value reflecting data characteristics in each time period may include any one of the following items: an average value of rotational speed data related to the predetermined component in this time period, a The median value of the rotational speed data related to the predetermined component, and the effective value of the rotational speed data related to the predetermined component within the time period.

Optionally, the time-frequency conversion module can be used to determine whether the rotational speed data related to the predetermined component in each time period is within a set rotational speed range, if the rotational speed related to the predetermined component in any time period If the data is within the set rotational speed range, frequency domain conversion is performed on the vibration acceleration data of the predetermined component in the any period of time to obtain a corresponding acceleration spectrum.

Optionally, for the time-frequency conversion module, it can be used to calculate the standard deviation or steady-state error of the rotational speed data related to the predetermined component in each time period, if the standard deviation or steady-state error calculated for any time period Within the set threshold range, it can be determined that the rotational speed data related to the predetermined component in the any time period is within the set rotational speed range, if the standard deviation or steady-state error calculated for any time period is not within the set speed range. Within the predetermined threshold range, it can be determined that the rotational speed data related to the predetermined component in any of the time periods is not within the set rotational speed range.

Optionally, the predetermined component may include any one of the following items: a wind turbine, a generator cogging, a gear box, and a rolling bearing, wherein, when the predetermined component is a wind turbine, the predetermined component The vibration acceleration data may be the vibration acceleration data of the nacelle of the wind turbine, the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine, and the type of abnormal vibration may include the fundamental wave frequency vibration abnormality and the fundamental wave of the wind turbine. The frequency double frequency vibration is abnormal; when the predetermined component is a generator cogging, the vibration acceleration data of the predetermined component can be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component can be wind The rotational speed data of the generator, the types of abnormal vibration may include abnormal vibration of the cogging frequency of the generator and abnormal vibration of the frequency multiplier of the cogging frequency of the generator; when the predetermined component is a gearbox, the vibration acceleration data of the predetermined component may be is the vibration acceleration data of the power gear or driven gear in the gearbox, the rotational speed data related to the predetermined component may be the rotational speed data of the shaft where the power gear or the driven gear is located, and the type of abnormal vibration may include the meshing frequency of the power gear and Abnormal frequency doubled vibration of the meshing frequency of the power gear, or abnormal frequency doubled vibration of the meshing frequency of the driven gear and the meshing frequency of the driven gear; when the predetermined component is a rolling bearing, the vibration acceleration data of the predetermined component may be the bearing of the rolling bearing The vibration acceleration data of the seat, the rotational speed data related to the predetermined component may be the rotational speed data of the rolling bearing, and the types of abnormal vibration may include abnormal vibration of the characteristic frequency of the rolling bearing fault and abnormal vibration of the double frequency of the characteristic frequency of the rolling bearing fault.

According to yet another aspect of an exemplary embodiment of the present invention, there is provided a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for identifying abnormal vibration.

According to yet another aspect of the exemplary embodiments of the present invention, there is provided a computing device, the computing device comprising: a processor; a memory storing a computer program, when the computer program is executed by the processor, the above-mentioned abnormal identification is realized Vibration method.

Using the method and device for identifying abnormal vibration according to the exemplary embodiments of the present invention, the predetermined component with abnormal vibration in the wind turbine can be located in time and accurately, which provides a strong support for effectively evaluating the vibration state of the predetermined component.

Description of drawings

The above and other objects, features and advantages of the exemplary embodiments of the present invention will become more apparent from the following detailed description in conjunction with the accompanying drawings which exemplarily illustrate the embodiments.

1 shows a flowchart of a method for identifying abnormal vibration according to an exemplary embodiment of the present invention;

FIG. 2 shows a flowchart of steps of determining whether a scatter point satisfies a predetermined linear distribution law according to an exemplary embodiment of the present invention;

3A to 3D respectively illustrate exemplary diagrams of rotational speed-frequency scattergrams according to an exemplary embodiment of the present invention;

4 shows a flow chart of the steps of determining whether a scatter point satisfies a predetermined linear distribution law based on a predetermined frequency linear model according to an exemplary embodiment of the present invention;

5 shows a flow chart of the steps of determining whether the scatter points satisfy a predetermined linear distribution law based on an objective function according to an exemplary embodiment of the present invention;

6 shows a structural diagram of a device for identifying abnormal vibration according to an exemplary embodiment of the present invention;

7 shows a structural diagram of an abnormal vibration analysis module according to an exemplary embodiment of the present invention;

FIG. 8 shows a structural diagram of a linear distribution determination module according to an exemplary embodiment of the present invention.

Detailed ways

Various example embodiments will now be described more fully with reference to the accompanying drawings, in which some example embodiments are shown.

FIG. 1 shows a flowchart of a method of identifying abnormal vibration according to an exemplary embodiment of the present invention.

Referring to FIG. 1 , in step S10 , operation data of predetermined components of the wind turbine in multiple time periods are acquired. Here, the operation data may include vibration acceleration data of a predetermined component and rotational speed data related to the predetermined component.

Preferably, the vibration acceleration data of the predetermined component may include vibration acceleration data in a first predetermined direction and vibration acceleration data in a second predetermined direction. As an example, the first predetermined direction may refer to the direction from the head to the tail of the wind turbine, and the second predetermined direction may refer to the direction perpendicular to the wind direction (eg, the field crew is standing downwind, facing the nose, the field crew's The left-right direction may be defined as a second predetermined direction).

In an exemplary embodiment of the present invention, it can be determined whether there is abnormal vibration of the predetermined component in the first predetermined direction by processing the vibration acceleration data in the first predetermined direction and the vibration acceleration data in the second predetermined direction respectively, or whether the predetermined component has abnormal vibration in the first predetermined direction or not. Whether there is abnormal vibration of the component in the second predetermined direction.

Here, the types of the abnormal vibration may include abnormal vibration of the fundamental frequency of the predetermined part and abnormal vibration of the double frequency of the predetermined part. As an example, the predetermined component may comprise any one of the following: a wind turbine, a generator cogging, a gearbox, a rolling bearing.

In a first embodiment, the predetermined component may be a wind turbine of a wind turbine. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine. Correspondingly, the types of abnormal vibrations may include the fundamental frequency vibration abnormality of the wind turbine and the multiplication frequency vibration abnormality of the fundamental frequency.

In a second embodiment, the predetermined component may be a generator cogging (eg, a generator stator cogging). In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine. Correspondingly, the type of abnormal vibration may include abnormal vibration of generator cogging frequency and/or abnormal vibration of multiple frequency of generator cogging frequency. Here, since the generator cogging frequency is equal to the number of stator cogs of the wind turbine × the rotational speed of the wind generator/60, and the number of stator cogs of the wind generator is greater than the number of pole pairs of the wind generator, therefore, the generator cogging frequency (or multiplier) is greater than the fundamental frequency (or multiplier) of the wind turbine. That is to say, whether there is abnormal vibration in the wind turbine cogging (whether there is a generator cogging or not) can be determined by analyzing the data of the high frequency part (the part corresponding to the generator cogging frequency and/or frequency doubling) in each acceleration spectrum frequency vibration anomalies and/or frequency-octave vibration anomalies of the generator cogging frequency), determine the wind power by analyzing the data of the low-frequency part of each acceleration spectrum (the part corresponding to the fundamental frequency and/or the multiplier frequency of the wind turbine) Whether there is abnormal vibration of the generator (whether there is abnormal vibration of the fundamental wave frequency of the wind turbine and/or abnormal vibration of the frequency multiplier of the fundamental frequency).

In the third embodiment, the predetermined component may be a gear box, and the gear box includes a power gear and a driven gear. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the power gear in the gearbox or the vibration acceleration data of the driven gear, and the rotational speed data related to the predetermined component may be the rotational speed data of the shaft where the power gear is located or from The speed data of the axis where the moving gear is located. Correspondingly, the types of abnormal vibration may include abnormal vibration of power gear meshing frequency, abnormal vibration of multiple frequency of power gear meshing frequency, abnormal vibration of driven gear meshing frequency and/or abnormal vibration of multiple frequency of driven gear meshing frequency. Here, the meshing frequency of the power gear (or driven gear) is equal to the number of teeth of the power gear (or the number of teeth of the driven gear)×the rotational speed of the shaft where the power gear (or driven gear) is located/60.

In the fourth embodiment, the predetermined component may be a rolling bearing, which here may refer to any rolling bearing among a plurality of rolling bearings in the wind turbine. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the bearing seat of the rolling bearing, and the rotational speed data related to the predetermined component may be the rotational speed data of the rolling bearing.

Correspondingly, the type of abnormal vibration may include abnormal vibration of the characteristic frequency of the rolling bearing fault and/or abnormal vibration of a multiplier of the characteristic frequency of the fault of the rolling bearing. Here, it should be understood that the rolling bearing may include a bearing inner ring, an outer ring, rolling elements, and a cage, and correspondingly, the predetermined components may refer to the bearing inner ring, outer ring, rolling elements or cage. For example, taking a rolling element whose predetermined component is a rolling bearing as an example, at this time, based on the method for identifying abnormal vibration of the exemplary embodiment of the present invention, it can be identified whether the rolling element has abnormal vibration of the fault characteristic frequency of the rolling element or the frequency multiplication of the fault characteristic frequency. Abnormal vibration. Here, the inner ring, outer ring, rolling body, and cage of the bearing correspond to their respective fault coefficients, and the fault characteristic frequency can be obtained by multiplying the fault coefficient and the rotational frequency.

In step S20, frequency domain conversion is performed on the vibration acceleration data of the predetermined component in multiple time periods, respectively, so as to obtain multiple corresponding acceleration spectra.

For example, the vibration acceleration data of the predetermined component in any one of the multiple time periods can be converted in the frequency domain to obtain the acceleration spectrum corresponding to the vibration acceleration data in the any one time period, that is, a time period The vibration acceleration data inside corresponds to an acceleration spectrum. As an example, the vibration acceleration data of the predetermined component in multiple time periods can be converted in the frequency domain through fast Fourier transform, however, the present invention is not limited to this, and other methods can also be used to perform the frequency domain conversion.

Preferably, based on the rotational speed data related to the predetermined component, the obtained operation data of the predetermined component in multiple time periods may be screened in advance, and the vibration acceleration data of the predetermined component in the filtered operation data in multiple time periods may be filtered. The frequency domain conversion is performed separately.

For example, it can be determined whether the rotational speed data related to the predetermined component in each time period is within the set rotational speed range, and if the rotational speed data related to the predetermined component in any time period is within the set rotational speed range, then the predetermined component is within the specified rotational speed range. Frequency domain conversion is performed on the vibration acceleration data in any of the above-mentioned time periods to obtain a corresponding acceleration spectrum. Here, since the rotational speed data in the wind turbine is a time variable, the fundamental frequency or frequency multiplication of predetermined components is related to the rotational speed, and it is difficult to complete the abnormal vibration identification of the fundamental frequency or frequency multiplication when the rotational speed fluctuates greatly. Therefore, in order to improve the abnormal vibration identification accuracy of the fundamental frequency or frequency multiplication, the fluctuation range of the rotational speed data can be limited, that is, the rotational speed fluctuation of the operating data used for abnormal vibration identification is small.

As an example, the step of determining whether the rotational speed data associated with the predetermined component within each time period is within a set rotational speed range may include calculating a standard deviation or steady-state error of rotational speed data associated with the predetermined component within each time period, if If the standard deviation or steady-state error calculated for any time period is within the set threshold range, it is determined that the rotational speed data related to the predetermined component in the any time period is within the set rotational speed range, if for any time period If the standard deviation or steady-state error calculated in the segment is not within the set threshold range, it is determined that the rotational speed data related to the predetermined component in any of the time periods is not within the set rotational speed range. Here, the set threshold range can be set according to the experience of those skilled in the art. Too wide setting of the threshold range will reduce the accuracy of abnormal vibration identification, and setting the threshold range too narrow will reduce the amount of time used for subsequent analysis and calculation. The amount of rotational speed data may even result in the absence of rotational speed data for analysis and calculation. Therefore, the setting of the threshold range should be able to constrain the rotational speed within a certain range, and ensure that it has sufficient use after screening. The speed data calculated in the subsequent analysis.

In step S30, a frequency value for abnormal vibration analysis in each acceleration spectrum is determined.

Here, the abscissa of each acceleration spectrum obtained by conversion can be the frequency value, and the ordinate can be the frequency amplitude value. In the exemplary embodiment of the present invention, the frequency amplitude value can be selected based on the comparison of the frequency amplitude value and the frequency amplitude threshold value for abnormal vibration. The analyzed frequency value.

In one case, the step of determining the frequency value used for abnormal vibration analysis in each acceleration spectrum may include: searching for a frequency point in the acceleration spectrum whose frequency amplitude value is greater than a frequency amplitude threshold value, and assigning the frequency value corresponding to the found frequency point. As the frequency value used for abnormal vibration analysis in this acceleration spectrum. At this time, one or more frequency values for abnormal vibration analysis can be determined from an acceleration spectrum.

In another case, the step of determining the frequency value used for abnormal vibration analysis in each acceleration spectrum may include: determining whether the frequency amplitude value corresponding to the preset frequency point of interest in the acceleration spectrum is greater than the frequency amplitude threshold, and if the frequency value is greater than the preset frequency amplitude threshold If the frequency amplitude value corresponding to the frequency of interest is greater than the frequency amplitude threshold, the frequency value corresponding to the preset frequency of interest is used as the frequency value in the acceleration spectrum for abnormal vibration analysis. The above determination method only judges the preset frequency points of interest, which is more accurate and more efficient than the first method of determining the frequency value for abnormal vibration analysis.

As an example, the preset frequency point of interest may be a predetermined number of frequency points before all the frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value. For example, it can generally be considered that the frequency value corresponding to the frequency point with the largest frequency amplitude value in the acceleration spectrum is the fundamental frequency of the predetermined component, and the frequency value corresponding to the frequency point with the second largest frequency amplitude value is twice the frequency of the predetermined component, and so on. . At this time, analyzing the preset frequency point of interest is equivalent to analyzing the fundamental frequency and its multiplier of the predetermined component, which can improve the accuracy of abnormal vibration identification.

In step S40, it is determined whether the predetermined component has abnormal vibration based on the determined frequency value for abnormal vibration analysis in each acceleration spectrum and rotational speed data related to the predetermined component in each time period.

Specifically, it can be determined whether a predetermined linear distribution law is satisfied between the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed data related to a predetermined component in each time period, and when the frequency value is related to the rotational speed When the predetermined linear distribution law is satisfied between the data, it is determined that the predetermined component has abnormal vibration. When the predetermined linear distribution law is not satisfied between the frequency value and the rotational speed data, it is determined that there is no abnormal vibration of the predetermined component. Here, the predetermined linear distribution law may be a distribution law for reflecting the linear relationship between the rotational speed and the frequency of the predetermined component.

Preferably, the step of determining whether a predetermined linear distribution law is satisfied between the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed data related to the predetermined component in each time period may include: The rotational speed data related to the predetermined component calculates the rotational speed statistical value reflecting the data characteristics in each time period, and determines whether a predetermined linear distribution law is satisfied between the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value.

的比值。As an example, the rotational speed statistics reflecting the data characteristics for each time period may include any one of the following items: an average value of rotational speed data related to the predetermined component in this time period, a The median value of rotational speed data related to the component, and the effective value of rotational speed data related to the predetermined component within the time period. As an example, the effective value of the rotational speed data may refer to the maximum value of the rotational speed data related to the predetermined component in the time period and the

ratio.

2, the steps of determining whether the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value satisfy a predetermined linear distribution law will be described by taking the rotational speed data related to the predetermined component as the rotational speed statistical value as an example.

FIG. 2 shows a flowchart of the steps of determining whether a scatter point satisfies a predetermined linear distribution law according to an exemplary embodiment of the present invention.

In step S201, a rotational speed-frequency scatter diagram is drawn based on the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value in each time period. Here, one scatter point in the rotational speed-frequency scattergram may correspond to the rotational speed statistical value of a time period and a frequency value used for abnormal vibration analysis in the acceleration spectrum corresponding to the time period.

In step S202, scatter points within a preset range of a predetermined frequency linear model are selected.

Preferably, all the scatter points included in the rotational speed-frequency scatter diagram can be screened, that is, the scatter points within the preset range of the predetermined frequency linear model can be selected for subsequent abnormal vibration analysis. Here, those skilled in the art can define the size of the preset range according to actual needs, and using scatter points within the preset range of the predetermined frequency linear model to perform abnormal vibration analysis can improve the accuracy of identification.

For example, the predetermined frequency linear model may be a model for abnormal vibration analysis of a predetermined component, that is, the predetermined frequency linear model may be a model capable of representing a linear relationship between the rotational speed and frequency of the predetermined component. Taking the predetermined component as a wind turbine as an example, the predetermined frequency linear model may be a model for reflecting the linear relationship between the fundamental wave frequency/multiple frequency of the wind turbine and the rotational speed of the wind turbine, for example, the predetermined frequency linear The model can be expressed as f _n =n×p×r/60, where f _n is the fundamental frequency of the wind turbine or the frequency multiple of the fundamental frequency, p is the number of pole pairs of the wind turbine, and r is the rotational speed of the wind turbine (rotor speed), n is an integer greater than or equal to 1. When n=1, f ₁ represents the fundamental frequency of the wind turbine, and when n≥2, f _n represents the frequency multiplication of the fundamental frequency of the wind turbine.

或

所包含的区域内。类似地，针对风力发电机基波频率的倍频(即，n≥2)的情况，预设范围可指被散点边界

或

所包含的区域内。For the case of the fundamental frequency of the wind turbine (ie, n=1), the preset range may refer to the boundary of the scattered points

or

within the included area. Similarly, for the case of the frequency multiplication of the fundamental frequency of the wind turbine (ie, n≥2), the preset range may refer to the boundary of the scattered points

or

within the included area.

As an example, the value ranges of the parameter b and the parameter k in the scatter boundary can be determined based on the grid-connected rotational speed range, the set rotational speed range (or the set threshold range) and the number of pole pairs of the wind turbine corresponding to the wind turbine. .

In step S203, based on the selected scatter points, it is determined whether the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value in each time period satisfy the predetermined linear distribution law of the predetermined frequency linear model. That is, the predetermined linear distribution law can be determined by the frequency linear model, and when the predetermined linear distribution law of the predetermined frequency linear model is satisfied between the frequency value and the rotational speed statistical value, it is determined that the predetermined component exists and the predetermined frequency linear model exists. The corresponding abnormal vibration.

For example, when the predetermined frequency linear model is a model for reflecting the linear relationship between the fundamental frequency of the wind turbine and the rotational speed of the wind turbine, if the predetermined frequency is satisfied between the frequency value and the rotational speed statistical value According to the predetermined linear distribution law of the linear model, it is determined that there is abnormal vibration of the fundamental frequency of the wind turbine. When the predetermined frequency linear model is a model for reflecting the linear relationship between the frequency multiplication of the fundamental frequency of the wind turbine and the rotational speed of the wind turbine, if the predetermined frequency value and the rotational speed statistical value satisfy the predetermined frequency According to the predetermined linear distribution law of the frequency linear model, it is determined that the wind turbine has abnormal vibration of a frequency that is a multiple of the fundamental frequency.

It should be understood that the manner of determining whether a predetermined linear distribution law is satisfied between the frequency value and the rotational speed statistical value shown in FIG. 2 is only an example, and those skilled in the art may use other manners to determine.

3A to 3D respectively illustrate example graphs of rotational speed-frequency scattergrams according to an exemplary embodiment of the present invention.

3A and 3B respectively show the rotational speed-frequency scatter diagram when the predetermined component is a wind turbine and the vibration acceleration data of the predetermined component is the vibration acceleration data of the first predetermined direction and the second predetermined direction, and the abscissa is the rotational speed statistics value (such as average rotational speed), the ordinate is the frequency value, curve 1 represents the predetermined frequency linear model of the fundamental wave frequency of the wind turbine, and curve 2 represents the predetermined frequency linear model of the multiplier (double frequency) of the fundamental wave frequency of the wind turbine . Taking curve 1 as an example, when it is desired to identify whether there is abnormal vibration of the fundamental frequency of the wind turbine, the scatter points within a predetermined range around curve 1 can be selected, and the selected one can be determined by linear regression parameter estimation or determination of the objective function. Whether the scattered points conform to the predetermined linear distribution law of the predetermined frequency linear model, if the selected scattered points conform to the predetermined linear distribution law of the predetermined frequency linear model, it indicates that the wind turbine has abnormal vibration of the fundamental frequency.

3C and 3D respectively show the rotational speed-frequency scatter diagram when the predetermined component is a wind turbine and the vibration acceleration data of the predetermined component is the vibration acceleration data in the first predetermined direction or the second predetermined direction, and the abscissa is the rotational speed statistics value (such as average rotation speed), the ordinate is the frequency value, curve 1 represents the predetermined frequency linear model of the fundamental frequency of the wind turbine, curve 2 represents the predetermined frequency linear model of the 2 times the fundamental frequency of the wind turbine, and curve 3 represents A predetermined frequency linear model of 3 times the fundamental frequency of the wind turbine. The specific identification method of abnormal vibration is the same as the identification method shown in FIG. 3A and FIG. 3B , and the present invention will not repeat the details of this part.

Preferably, it can be determined by a predetermined frequency linear model for reflecting a predetermined linear distribution law or by an objective function for indicating the distance of each scatter point (scatter point formed by the frequency value and the rotational speed data) to the predetermined frequency linear model It is determined whether a predetermined linear distribution law is satisfied between the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed data related to the predetermined component in each time period.

The steps of determining whether the scatter points satisfy the predetermined linear distribution law based on the predetermined frequency linear model will be described below with reference to FIG. 4 .

FIG. 4 shows a flowchart of the steps of determining whether the scatter points satisfy a predetermined linear distribution law based on a predetermined frequency linear model according to an exemplary embodiment of the present invention.

As shown in FIG. 4 , in step S401 , linear regression is performed on the frequency value corresponding to the selected scatter point and the rotational speed data to obtain model parameters of the linear regression. As an example, the least squares method or the maximum likelihood method can be used to perform linear regression on the frequency value and rotational speed data corresponding to the scattered points to obtain a linear regression model, and then determine the model parameters of the linear regression model. However, the present invention is not limited to this, and other methods can also be used to perform linear regression analysis on the scattered points.

In step S402, the difference between the model parameter and the specified parameter of the predetermined component is calculated, and it is determined whether the difference is greater than the first set value.

As an example, when the predetermined component is a wind turbine, the specified parameter may be the number of magnetic pole pairs of the wind turbine; when the predetermined component is a generator cogging, the specified parameter may be the number of magnetic pole pairs of the wind turbine; When the component is a gearbox, the specified parameter can be the number of teeth of the power gear or the number of teeth of the driven gear. When the predetermined component is a rolling bearing, the specified parameter can be the failure factor of the rolling bearing (such as the inner ring, outer ring, rolling bearing of the rolling bearing). failure factor corresponding to one of the body and the cage).

再将模型参数

与风力发电机的磁极对数p作差。For example, taking the predetermined component as a wind turbine as an example, perform linear regression on the frequency value and rotational speed data corresponding to the selected scatter points to obtain the model parameters

Then the model parameters

Difference with the number of pole pairs p of the wind turbine.

If the difference between the model parameter and the specified parameter of the predetermined component is not greater than the first set value, step S403 is executed: it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data satisfy the predetermined frequency linear model. A predetermined linear distribution law.

If the difference between the model parameter and the specified parameter of the predetermined component is greater than the first set value, step S404 is executed: it is determined that the predetermined frequency linearity is not satisfied between the frequency value corresponding to the selected scatter point and the rotational speed data The predetermined linear distribution law of the model.

The following describes the steps of determining whether the scatter points satisfy the predetermined linear distribution law based on the objective function with reference to FIG. 5 .

FIG. 5 shows a flowchart of the steps of determining whether the scatter points satisfy a predetermined linear distribution law based on an objective function according to an exemplary embodiment of the present invention.

As shown in FIG. 5, in step S501, an objective function is established. Here, the objective function may indicate the distance of each scatter point to a predetermined frequency linear model. Preferably, the objective function may indicate the root mean square value of the distance of each scatter point to a predetermined frequency linear model.

For example, the objective function can be expressed by the following formula:

In formula (1), y represents the objective function, N is the number of scatter points, f _i ′ is the ordinate value of the i-th scatter point in the speed-frequency scatter diagram, and _{ri ′ is the corresponding value of f i ′} . The abscissa value in the rotational speed-frequency scatter plot, that is, the rotational speed statistical value corresponding to the i-th scatter point.

In step S502, the value of the objective function is obtained by substituting the frequency value and rotational speed data corresponding to the selected scatter points into the objective function. For example, the value of the objective function y can be calculated by the above formula (1).

In step S503, it is determined whether the value of the objective function is greater than the second set value.

If the value of the objective function is not greater than the second set value, step S504 is performed: it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data satisfy a predetermined linear distribution law of a predetermined frequency linear model.

If the value of the objective function is greater than the second set value, step S505 is executed: it is determined that the frequency value corresponding to the selected scatter point and the rotational speed data do not satisfy the predetermined linear distribution law of the predetermined frequency linear model.

Preferably, the value ranges of the first set value and the second set value may be determined based on the scatter boundary (ie, the preset range of the predetermined frequency linear model).

FIG. 6 shows a structural diagram of an apparatus for identifying abnormal vibration according to an exemplary embodiment of the present invention.

As shown in FIG. 6 , the apparatus for identifying abnormal vibration according to an exemplary embodiment of the present invention includes an operation data acquisition module 10 , a time-frequency conversion module 20 , a frequency value determination module 30 and an abnormal vibration analysis module 40 .

Specifically, the operation data acquisition module 10 is configured to acquire operation data of predetermined components of the wind turbine in multiple time periods. Here, the operation data may include vibration acceleration data of a predetermined component and rotational speed data related to the predetermined component.

Preferably, the vibration acceleration data of the predetermined component may include vibration acceleration data in a first predetermined direction and vibration acceleration data in a second predetermined direction. As an example, the first predetermined direction may refer to the direction from the head to the tail of the wind turbine, and the second predetermined direction may refer to the direction perpendicular to the wind direction (eg, the field crew is standing downwind, facing the nose, the field crew's The left-right direction may be defined as a second predetermined direction).

In an exemplary embodiment of the present invention, it can be determined whether there is abnormal vibration of the predetermined component in the first predetermined direction by processing the vibration acceleration data in the first predetermined direction and the vibration acceleration data in the second predetermined direction respectively, or whether the predetermined component has abnormal vibration in the first predetermined direction or not. Whether there is abnormal vibration of the component in the second predetermined direction.

Here, the types of the abnormal vibration may include abnormal vibration of the fundamental frequency of the predetermined part and abnormal vibration of the double frequency of the predetermined part. As an example, the predetermined component may comprise any one of the following: a wind turbine, a generator cogging, a gearbox, a rolling bearing.

In a first embodiment, the predetermined component may be a wind turbine of a wind turbine. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine. Correspondingly, the types of abnormal vibrations may include the fundamental frequency vibration abnormality of the wind turbine and the multiplication frequency vibration abnormality of the fundamental frequency.

In a second embodiment, the predetermined component may be a generator cogging (eg, a generator stator cogging). In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the wind turbine nacelle, and the rotational speed data related to the predetermined component may be the rotational speed data of the wind turbine. Correspondingly, the type of abnormal vibration may include abnormal vibration of generator cogging frequency and/or abnormal vibration of multiple frequency of generator cogging frequency. Here, since the generator cogging frequency is equal to the number of stator cogs of the wind turbine × the rotational speed of the wind generator/60, and the number of stator cogs of the wind generator is greater than the number of pole pairs of the wind generator, therefore, the generator cogging frequency (or multiplier) is greater than the fundamental frequency (or multiplier) of the wind turbine. That is to say, whether there is abnormal vibration in the wind turbine cogging (whether there is a generator cogging or not) can be determined by analyzing the data of the high frequency part (the part corresponding to the generator cogging frequency and/or frequency doubling) in each acceleration spectrum frequency vibration anomalies and/or frequency-octave vibration anomalies of the generator cogging frequency), determine the wind power by analyzing the data of the low-frequency part of each acceleration spectrum (the part corresponding to the fundamental frequency and/or the multiplier frequency of the wind turbine) Whether there is abnormal vibration of the generator (whether there is abnormal vibration of the fundamental wave frequency of the wind turbine and/or abnormal vibration of the frequency multiplier of the fundamental frequency).

In the third embodiment, the predetermined component may be a gear box, and the gear box includes a power gear and a driven gear. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the power gear in the gearbox or the vibration acceleration data of the driven gear, and the rotational speed data related to the predetermined component may be the rotational speed data of the shaft where the power gear is located or from The speed data of the axis where the moving gear is located. Accordingly, the types of abnormal vibration may include abnormal vibration of power gear meshing frequency, abnormal vibration of multiplier of meshing frequency of power gear, abnormal vibration of driven gear meshing frequency, and abnormal vibration of multiplier of meshing frequency of driven gear. Here, the meshing frequency of the power gear (or driven gear) is equal to the number of teeth of the power gear (or the number of teeth of the driven gear)×the rotational speed of the shaft where the power gear (or driven gear) is located/60.

In the fourth embodiment, the predetermined component may be a rolling bearing, which here may refer to any rolling bearing among a plurality of rolling bearings in the wind turbine. In this case, the vibration acceleration data of the predetermined component may be the vibration acceleration data of the bearing seat of the rolling bearing, and the rotational speed data related to the predetermined component may be the rotational speed data of the rolling bearing. Correspondingly, the type of abnormal vibration may include abnormal vibration of the characteristic frequency of the rolling bearing fault and/or abnormal vibration of a multiplier of the characteristic frequency of the fault of the rolling bearing. Here, it should be understood that the rolling bearing may include a bearing inner ring, an outer ring, rolling elements, and a cage, and correspondingly, the predetermined components may refer to the bearing inner ring, outer ring, rolling elements or cage. For example, taking a rolling element whose predetermined component is a rolling bearing as an example, it can be identified whether the rolling element has abnormal vibration of the fault characteristic frequency of the rolling element or abnormal vibration of the double frequency of the fault characteristic frequency of the rolling element. Here, the inner ring, outer ring, rolling body, and cage of the bearing correspond to their respective fault coefficients, and the fault characteristic frequency can be obtained by multiplying the fault coefficient and the rotational frequency.

The time-frequency conversion module 20 is configured to perform frequency domain conversion on the vibration acceleration data of the predetermined component in multiple time periods respectively, so as to obtain multiple corresponding acceleration spectra.

For example, the time-frequency conversion module 20 can be configured to perform frequency domain conversion on the vibration acceleration data of the predetermined component in any time period among the multiple time periods to obtain the vibration acceleration data corresponding to the vibration acceleration data in the any time period. The acceleration spectrum, that is, the vibration acceleration data in a time period corresponds to an acceleration spectrum. As an example, the time-frequency conversion module 20 can perform frequency domain conversion on the vibration acceleration data of the predetermined component in multiple time periods through fast Fourier transform, but the present invention is not limited to this, and other methods can also be used to perform frequency domain conversion .

Preferably, the time-frequency conversion module 20 can be configured to pre-screen the acquired operating data of the predetermined components in multiple time periods based on the rotational speed data related to the predetermined components, and select the predetermined components in the filtered operating data in the multiple time periods. The vibration acceleration data in each time period are converted into frequency domain.

For example, the time-frequency conversion module 20 can be used to determine whether the rotational speed data related to the predetermined component in each time period is within the set rotational speed range, if the rotational speed data related to the predetermined component in any time period is within Within the set rotational speed range, frequency domain conversion is performed on the vibration acceleration data of the predetermined component in any of the time periods, so as to obtain a corresponding acceleration spectrum.

As an example, the time-frequency conversion module 20 can be used to calculate the standard deviation or steady-state error of the rotational speed data related to the predetermined component in each time period, if the calculated standard deviation or steady-state error for any time period is in the set Within the threshold range, it is determined that the rotational speed data related to the predetermined component in any time period is within the set rotational speed range, if the standard deviation or steady-state error calculated for any time period is not within the set threshold range , then it is determined that the rotational speed data related to the predetermined component in any of the time periods is not within the set rotational speed range. Here, the set threshold range can be set according to the experience of those skilled in the art.

The frequency value determination module 30 is used for determining the frequency value in each acceleration spectrum for abnormal vibration analysis.

Here, the abscissa of each acceleration spectrum obtained by conversion may be the frequency value, and the ordinate may be the frequency amplitude value. In an exemplary embodiment of the present invention, the frequency value determination module 30 may determine the frequency value based on the comparison between the frequency amplitude value and the frequency amplitude threshold. Select the frequency value to use for abnormal vibration analysis.

In one case, the frequency value determination module 30 can be used to find a frequency point whose frequency amplitude value is greater than the frequency amplitude threshold value in the acceleration spectrum, and use the frequency value corresponding to the found frequency point as the frequency value in the acceleration spectrum for abnormal vibration analysis. frequency value. At this time, one or more frequency values for abnormal vibration analysis can be determined from an acceleration spectrum.

In another case, the frequency value determination module 30 can be used to determine whether the frequency amplitude value corresponding to the preset frequency of interest in the acceleration spectrum is greater than the frequency amplitude threshold, and if the frequency amplitude value corresponding to the preset frequency of interest is greater than the frequency If the amplitude threshold is set, the frequency value corresponding to the preset frequency point of interest is used as the frequency value in the acceleration spectrum for abnormal vibration analysis.

As an example, the preset frequency point of interest may be a predetermined number of frequency points before all the frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value. For example, it can generally be considered that the frequency value corresponding to the frequency point with the largest frequency amplitude value in the acceleration spectrum is the fundamental frequency of the predetermined component, and the frequency value corresponding to the frequency point with the second largest frequency amplitude value is twice the frequency of the predetermined component, and so on. . In this case, analyzing the frequency point of interest is equivalent to analyzing the fundamental frequency and its multiplier of the predetermined component, which can improve the accuracy of abnormal vibration identification.

The abnormal vibration analysis module 40 is configured to determine whether the predetermined component has abnormal vibration based on the determined frequency value for abnormal vibration analysis in each acceleration spectrum and rotational speed data related to the predetermined component in each time period.

Specifically, the abnormal vibration analysis module 40 is configured to determine whether a predetermined linear distribution law is satisfied between the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed data related to the predetermined component in each time period. When a predetermined linear distribution law is satisfied between the frequency value and the rotational speed data, it is determined that the predetermined component has abnormal vibration. When the predetermined linear distribution law is not satisfied between the frequency value and the rotational speed data, it is determined that there is no abnormal vibration of the predetermined component. Here, the predetermined linear distribution law may be a distribution law for reflecting the linear relationship between the rotational speed and the frequency of the predetermined component.

Preferably, the abnormal vibration analysis module 40 can be configured to calculate the rotational speed statistics value reflecting the data characteristics in each time period based on the rotational speed data related to the predetermined component in each time period, and determine the abnormal vibration analysis in each acceleration spectrum for abnormal vibration analysis. Whether the predetermined linear distribution law is satisfied between the frequency value and the rotational speed statistical value.

的比值。As an example, the rotational speed statistics reflecting the data characteristics for each time period may include any one of the following items: an average value of rotational speed data related to the predetermined component in this time period, a The median value of rotational speed data related to the component, and the effective value of rotational speed data related to the predetermined component within the time period. Here, the effective value of the rotational speed data may refer to the maximum value of the rotational speed data related to the predetermined component within the time period and

ratio.

7 , the process of determining whether the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value satisfies a predetermined linear distribution law is described by taking the rotational speed data related to the predetermined component as the rotational speed statistical value as an example.

FIG. 7 shows a structural diagram of an abnormal vibration analysis module according to an exemplary embodiment of the present invention.

As shown in FIG. 7 , the abnormal vibration analysis module 40 according to an exemplary embodiment of the present invention may include a scatter plot drawing module 41 , a scatter screening module 42 and a linear distribution determination module 43 .

Specifically, the scatter diagram drawing module 41 is configured to draw a rotational speed-frequency scatter diagram based on the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value in each time period. Here, one scatter point in the rotational speed-frequency scattergram may correspond to the rotational speed statistical value of a time period and a frequency value used for abnormal vibration analysis in the acceleration spectrum corresponding to the time period.

The scatter point screening module 42 is used for selecting scatter points within a preset range of a predetermined frequency linear model.

Preferably, the scatter screening module 42 can be used to screen all the scatter points included in the rotational speed-frequency scatter diagram, that is, to select the scatter points within the preset range of the predetermined frequency linear model for subsequent abnormal vibration analysis . Here, those skilled in the art can define the size of the preset range according to actual needs, and using scatter points within the preset range of the predetermined frequency linear model to perform abnormal vibration analysis can improve the accuracy of identification.

The linear distribution determination module 43 is configured to determine, based on the selected scatter points, whether the frequency value used for abnormal vibration analysis in each acceleration spectrum and the rotational speed statistical value in each time period satisfy the predetermined linear distribution of the predetermined frequency linear model law. That is to say, the predetermined linear distribution law can be determined by the frequency linear model. When the predetermined linear distribution law of the predetermined frequency linear model is satisfied between the frequency value and the rotational speed statistical value, the linear distribution determining module 43 is used to determine The predetermined component has abnormal vibration corresponding to the predetermined frequency linear model.

The following describes the process of determining whether the scattered points satisfy the predetermined linear distribution law based on the predetermined frequency linear model.

The linear distribution determination module 43 is configured to perform linear regression on the frequency value corresponding to the selected scatter point and the rotational speed data, obtain the model parameters of the linear regression, calculate the difference between the model parameters and the specified parameters of the predetermined component, and Determine whether the difference is greater than the first set value, and if the difference between the model parameter and the specified parameter of the predetermined component is not greater than the first set value, then determine the frequency value and the rotational speed corresponding to the selected scatter point The data satisfy the predetermined linear distribution law of the predetermined frequency linear model. If the difference between the model parameter and the specified parameter of the predetermined component is greater than the first set value, it is determined that the predetermined linearity of the predetermined frequency linear model is not satisfied between the frequency value corresponding to the selected scatter point and the rotational speed data Distribution.

As an example, when the predetermined component is a wind turbine, the specified parameter may be the number of magnetic pole pairs of the wind turbine; when the predetermined component is a generator cogging, the specified parameter may be the number of magnetic pole pairs of the wind turbine; When the component is a gearbox, the specified parameter can be the number of teeth of the power gear or the number of teeth of the driven gear. When the predetermined component is a rolling bearing, the specified parameter can be the failure factor of the rolling bearing (such as the inner ring, outer ring, rolling bearing of the rolling bearing). failure factor corresponding to one of the body and the cage).

The following describes the process of determining whether the scatter points satisfy the predetermined linear distribution law based on the objective function with reference to FIG. 8 .

FIG. 8 shows a structural diagram of a linear distribution determination module according to an exemplary embodiment of the present invention.

As shown in FIG. 8 , the linear distribution determination module 43 according to an exemplary embodiment of the present invention may include an objective function establishment submodule 431 , an objective function value calculation submodule 432 and a distribution law determination submodule 433 .

Specifically, the objective function establishment sub-module 431 is used to establish the objective function. Here, the objective function may indicate the distance of each scatter point to a predetermined frequency linear model. Preferably, the objective function may indicate the root mean square value of the distance of each scatter point to a predetermined frequency linear model.

The objective function value calculation sub-module 432 is configured to obtain the value of the objective function by substituting the frequency value and rotational speed data corresponding to the selected scatter points into the objective function.

The distribution law determination sub-module 433 is used to determine whether the value of the objective function is greater than the second set value. If the value of the objective function is not greater than the second set value, the distribution law determination sub-module 433 determines a predetermined linear distribution law that satisfies a predetermined frequency linear model between the frequency value corresponding to the selected scatter point and the rotational speed data. If the value of the objective function is greater than the second set value, the distribution law determination sub-module 433 is configured to determine that the frequency value corresponding to the selected scatter point and the rotational speed data do not satisfy the predetermined linearity of the predetermined frequency linear model. Distribution. Preferably, the value ranges of the first set value and the second set value may be determined based on the scatter boundary (ie, the preset range of the predetermined frequency linear model).

A computing device is also provided according to an exemplary embodiment of the present invention. The computing device includes a processor and memory. Memory is used to store computer programs. The computer program is executed by the processor to cause the processor to execute the computer program of the method of identifying abnormal vibration as described above.

Exemplary embodiments according to the present invention also provide a computer-readable storage medium storing a computer program. The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the above-described method of identifying abnormal vibration. The computer-readable recording medium is any data storage device that can store data read by a computer system. Examples of the computer-readable recording medium include read-only memory, random-access memory, optical disks, magnetic tapes, floppy disks, optical data storage devices, and carrier waves (such as data transmission through the Internet via wired or wireless transmission paths).

Using the method and device for identifying abnormal vibration according to the exemplary embodiment of the present invention, the components with abnormal frequency vibration in the wind turbine can be located timely and accurately, which provides strong support for quickly and effectively evaluating the vibration state of predetermined components. In addition, it can effectively improve the efficiency of unit fault location and save operation and maintenance costs.

Although the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that form and detail may be made therein without departing from the spirit and scope of the invention as defined in the claims various changes on.

Claims (20)

1. A method of identifying fundamental frequency abnormal vibrations of a predetermined component of a wind park, the method comprising:

acquiring operation data of a preset component of a wind generating set in a plurality of time periods, wherein the operation data comprises vibration acceleration data of the preset component and rotating speed data related to the preset component;

respectively carrying out frequency domain conversion on the vibration acceleration data of the predetermined component in the plurality of time periods to obtain a plurality of acceleration frequency spectrums corresponding to each other;

determining a frequency value for abnormal vibration analysis in each acceleration frequency spectrum;

drawing a rotating speed-frequency scatter diagram based on the frequency values and the rotating speed data, wherein one scattering point in the rotating speed-frequency scatter diagram corresponds to the rotating speed data of a time period and one frequency value used for abnormal vibration analysis in an acceleration frequency spectrum corresponding to the time period;

selecting scattered points within a preset range of a linear model with a preset frequency;

performing linear regression on the frequency value corresponding to the selected scattering point and the rotating speed data to obtain a model parameter of the linear regression;

calculating a difference between the model parameter and a specified parameter of the predetermined component;

when the difference value is not larger than a first set value, determining that the frequency value corresponding to the selected scattering point and the rotating speed data meet a preset linear distribution rule of a preset frequency linear model;

the predetermined frequency linear model is a model for reflecting a linear relation between a fundamental frequency of the wind power generator and a rotation speed of the wind power generator, and when it is determined that the predetermined linear distribution rule is satisfied, it is determined that abnormal vibration of the fundamental frequency exists in the predetermined component, wherein the predetermined component includes a generator tooth groove, a gear box or a rolling bearing.

2. The method of claim 1, wherein the step of determining a frequency value for abnormal vibration analysis in each acceleration spectrum comprises:

searching for a frequency point of which the frequency amplitude value is greater than a frequency amplitude threshold value in the acceleration frequency spectrum;

and taking the frequency value corresponding to the searched frequency point as the frequency value for analyzing the abnormal vibration in the acceleration frequency spectrum.

3. The method of claim 1, wherein the step of determining a frequency value for abnormal vibration analysis in each acceleration spectrum comprises:

determining whether a frequency amplitude value corresponding to a preset concerned frequency point in the acceleration frequency spectrum is greater than a frequency amplitude threshold value;

and if the frequency amplitude value corresponding to the preset attention frequency point is larger than a frequency amplitude threshold value, taking the frequency value corresponding to the preset attention frequency point as a frequency value for analyzing abnormal vibration in the acceleration frequency spectrum.

4. The method according to claim 3, wherein the predetermined frequency points of interest are a predetermined number of previous frequency points in which all frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value.

5. The method of claim 1, wherein the method further comprises:

calculating a rotation speed statistic value reflecting data characteristics for each period of time based on the rotation speed data to identify abnormal vibration of the fundamental wave frequency of the predetermined component based on the rotation speed statistic value.

6. The method of claim 5, wherein the data-characteristic-reflecting rotational speed statistics for each time period comprise any one of: an average value of the rotational speed data related to the predetermined component during the time period, a median value of the rotational speed data related to the predetermined component during the time period, and an effective value of the rotational speed data related to the predetermined component during the time period.

7. The method of claim 1, wherein the step of frequency-domain converting the vibration acceleration data of the predetermined component over the plurality of time periods to obtain a corresponding plurality of acceleration spectra comprises:

determining whether the rotating speed data related to the preset component in each time period is in a set rotating speed range;

and if the rotating speed data related to the preset component in any time period is in a set rotating speed range, performing frequency domain conversion on the vibration acceleration data of the preset component in any time period to obtain a corresponding acceleration frequency spectrum.

8. The method of claim 7, wherein the step of determining whether the rotational speed data associated with the predetermined component for each time period is within a set rotational speed range comprises:

calculating a standard deviation or steady state error of the rotational speed data associated with the predetermined component for each time period;

determining that the rotational speed data associated with the predetermined component for any one time period is within a set rotational speed range if the standard deviation or steady state error calculated for the time period is within a set threshold range;

and if the standard deviation or the steady-state error calculated for any time period is not in the set threshold value range, determining that the rotating speed data related to the predetermined component in any time period is not in the set rotating speed range.

9. The method of claim 1,

when the preset component is a generator tooth space, the vibration acceleration data of the preset component is the vibration acceleration data of a cabin of the wind driven generator, the rotating speed data related to the preset component is the rotating speed data of the wind driven generator, and the type of abnormal vibration comprises generator tooth space frequency vibration abnormality and frequency doubling abnormal vibration of the generator tooth space frequency;

when the preset component is a gearbox, the vibration acceleration data of the preset component is vibration acceleration data of a power gear or a driven gear in the gearbox, the rotating speed data related to the preset component is rotating speed data of a shaft where the power gear or the driven gear is located, and the type of abnormal vibration comprises frequency multiplication abnormal vibration of the meshing frequency of the power gear and the meshing frequency of the power gear or frequency multiplication abnormal vibration of the meshing frequency of the driven gear and the meshing frequency of the driven gear;

when the preset component is a rolling bearing, the vibration acceleration data of the preset component is the vibration acceleration data of a bearing seat of the rolling bearing, the rotating speed data related to the preset component is the rotating speed data of the rolling bearing, and the type of abnormal vibration comprises rolling bearing fault characteristic frequency vibration abnormality and frequency doubling abnormal vibration of the rolling bearing fault characteristic frequency.

10. An apparatus for identifying fundamental frequency abnormal vibrations of a predetermined component of a wind turbine generator set, said apparatus comprising:

the operation data acquisition module is used for acquiring operation data of a preset component of the wind generating set in a plurality of time periods, wherein the operation data comprises vibration acceleration data of the preset component and rotating speed data related to the preset component;

the time-frequency conversion module is used for respectively carrying out frequency domain conversion on the vibration acceleration data of the preset component in the time periods so as to obtain a plurality of corresponding acceleration frequency spectrums;

the frequency value determining module is used for determining a frequency value used for abnormal vibration analysis in each acceleration frequency spectrum;

an abnormal vibration analysis module for determining whether there is abnormal vibration in the predetermined component based on the frequency value for abnormal vibration analysis in each acceleration frequency spectrum determined and the rotational speed data associated with the predetermined component for each period of time,

wherein the abnormal vibration analysis module includes:

the scatter diagram drawing module is used for drawing a rotating speed-frequency scatter diagram based on the frequency values and the rotating speed data, wherein one scatter point in the rotating speed-frequency scatter diagram corresponds to the rotating speed data of a time period and one frequency value used for abnormal vibration analysis in an acceleration frequency spectrum corresponding to the time period;

the scattered point screening module is used for selecting scattered points within a preset range of the linear model with the preset frequency;

a linear distribution determining module, configured to perform linear regression on the frequency value corresponding to the selected scattering point and the rotation speed data to obtain a linear regression model parameter, calculate a difference between the model parameter and a specified parameter of the predetermined component, and determine that a predetermined linear distribution rule of the predetermined frequency linear model is satisfied between the frequency value corresponding to the selected scattering point and the rotation speed data when the difference is not greater than a first set value,

the predetermined frequency linear model is a model for reflecting a linear relation between a fundamental frequency of the wind power generator and a rotation speed of the wind power generator, and when it is determined that the predetermined linear distribution rule is satisfied, it is determined that abnormal vibration of the fundamental frequency exists in the predetermined component, wherein the predetermined component includes a generator tooth groove, a gear box or a rolling bearing.

11. The apparatus according to claim 10, wherein the frequency value determining module is configured to search for a frequency point in the acceleration spectrum where the frequency amplitude value is greater than the frequency amplitude threshold value, and use a frequency value corresponding to the searched frequency point as the frequency value in the acceleration spectrum for abnormal vibration analysis.

12. The apparatus according to claim 10, wherein the frequency value determining module is configured to determine whether a frequency amplitude value corresponding to a preset frequency of interest in the acceleration spectrum is greater than a frequency amplitude threshold value, and if the frequency amplitude value corresponding to the preset frequency of interest is greater than the frequency amplitude threshold value, the frequency value determining module takes the frequency value corresponding to the preset frequency of interest as the frequency value for analyzing the abnormal vibration in the acceleration spectrum.

13. The apparatus according to claim 12, wherein the preset frequency points of interest are a predetermined number of previous frequency points in which all frequency points included in the acceleration spectrum are arranged in descending order according to the magnitude of the frequency amplitude value.

14. The apparatus according to claim 10, wherein the abnormal vibration analyzing module is configured to calculate a rotation speed statistic value reflecting a data characteristic for each period of time based on the rotation speed data to identify the abnormal vibration of the fundamental wave frequency of the predetermined component based on the rotation speed statistic value.

15. The apparatus of claim 14, wherein the data-characteristic-reflecting rotational speed statistics for each time period comprise any one of: an average value of the rotational speed data related to the predetermined component during the time period, a median value of the rotational speed data related to the predetermined component during the time period, and an effective value of the rotational speed data related to the predetermined component during the time period.

16. The apparatus of claim 10, wherein the time-frequency conversion module is configured to determine whether the rotation speed data associated with the predetermined component in each time period is within a set rotation speed range, and if the rotation speed data associated with the predetermined component in any time period is within the set rotation speed range, perform frequency domain conversion on the vibration acceleration data of the predetermined component in any time period to obtain a corresponding acceleration spectrum.

17. The apparatus of claim 16, wherein the time-frequency transform module is configured to calculate a standard deviation or a steady state error of the rotational speed data associated with the predetermined component for each time period, determine that the rotational speed data associated with the predetermined component for any time period is within a set rotational speed range if the standard deviation or the steady state error calculated for the time period is within a set threshold range,

and if the standard deviation or the steady-state error calculated for any time period is not in the set threshold value range, determining that the rotating speed data related to the predetermined component in any time period is not in the set rotating speed range.

18. The apparatus of claim 10,

when the preset component is a generator tooth space, the vibration acceleration data of the preset component is the vibration acceleration data of a cabin of the wind driven generator, the rotating speed data related to the preset component is the rotating speed data of the wind driven generator, and the type of abnormal vibration comprises generator tooth space frequency vibration abnormality and frequency doubling abnormal vibration of the generator tooth space frequency;

when the preset component is a gearbox, the vibration acceleration data of the preset component is vibration acceleration data of a power gear or a driven gear in the gearbox, the rotating speed data related to the preset component is rotating speed data of a shaft where the power gear or the driven gear is located, and the type of abnormal vibration comprises frequency multiplication abnormal vibration of the meshing frequency of the power gear and the meshing frequency of the power gear or frequency multiplication abnormal vibration of the meshing frequency of the driven gear and the meshing frequency of the driven gear;

when the preset component is a rolling bearing, the vibration acceleration data of the preset component is the vibration acceleration data of a bearing seat of the rolling bearing, the rotating speed data related to the preset component is the rotating speed data of the rolling bearing, and the type of abnormal vibration comprises rolling bearing fault characteristic frequency vibration abnormality and frequency doubling abnormal vibration of the rolling bearing fault characteristic frequency.

19. A computer-readable storage medium storing a computer program which, when executed by a processor, implements the method of identifying fundamental frequency abnormal vibrations of a predetermined component of a wind park as claimed in any one of claims 1 to 9.

20. A computing device, the computing device comprising:

a processor;

memory storing a computer program which, when executed by the processor, implements the method of identifying fundamental frequency abnormal vibrations of a predetermined component of a wind park according to any one of claims 1 to 9.

Images