JP6899109B2 - Abnormality diagnosis method of the part to be diagnosed in the rotation drive device and the abnormality diagnosis device used for it. - Google Patents

Description

The present invention relates to a method for diagnosing an abnormality of a diagnosis target portion in a rotary drive device and an abnormality diagnosing device used therein. Specifically, using the vibration characteristic data output from the sensor, it is possible to reliably detect abnormalities in the parts to be diagnosed, such as bearings, gears, and chains, especially in drive devices that rotate at low speeds. The purpose is to provide a method and an abnormality diagnostic device used for the method.

In the case of a drive device including a bearing that rotates at a low speed of 200 rpm or less, such as a transport device such as a bucket elevator, a rotary kiln, or a drive unit of a mining device, if an abnormality diagnosis is to be performed using the vibration characteristic data output from the sensor. It has been pointed out that the S (signal) / N (noise) ratio of the obtained signal is low, and it is not possible to accurately detect an abnormality.
Therefore, the conventional abnormality diagnosis is performed exclusively by the five senses of the inspector, and a high degree of skill is required to ensure its accuracy.

On the other hand, the following method is known as a means for removing noise from a signal having a low S / N ratio.
For example, there are a bearing diagnosis method using a high-pass filter as described in Patent Document 1 and a method for extracting an abnormal signal based on a power spectral ratio as described in Patent Document 2. However, the former method has a problem that it is difficult to determine the cutoff frequency and it is not possible to select and extract the abnormal signal, and it is difficult to extract the abnormal signal sufficiently accurately. Further, the latter method has a problem that it is not suitable for extracting an abnormal signal of equipment accompanied by shocking vibration, and it is difficult to extract an abnormal signal sufficiently accurately.

Japanese Unexamined Patent Publication No. 2001-033353 Japanese Patent No. 3920715

An object of the present invention is to use the vibration characteristic data output from the sensor in view of the above-mentioned conventional problems, and to use the vibration characteristic data, in particular, to detect an abnormality in a diagnosis target portion such as a bearing, a gear, or a chain in a drive device rotating at a low speed. An object of the present invention is to provide an abnormality diagnosis method capable of reliably performing detection and an abnormality diagnosis device used for the method.

As a result of diligent research to achieve the above object, the present inventors have conducted data in the frequency domain obtained by Fourier transform of a diagnostic signal in addition to processing by a conventional statistical information filter that selects and extracts anomalous signals. We found that the above objectives can be achieved by analyzing the data in the time domain using new feature parameters, analyzing the time domain data using specific feature parameters, and evaluating these results by principal component analysis. We have come to propose the present invention.

b)時間領域に係る特徴パラメータ

c)時間領域に係る特徴パラメータ

d)時間領域に係る特徴パラメータ

e) 診断対象部の低周波帯域の安定指数を表わす特徴パラメータ

f）診断対象部の低周波帯域のパワースペクトルの総和を表わす特徴パラメータ

g）診断対象部のパス周波数成分率を表わす特徴パラメータ

h）診断対象部の損傷指数を表わす特徴パラメータ

i）診断対象部のパス周波数２次高調波率を表わす特徴パラメータ

j）診断対象部のパス周波数３次高調波率を表わす特徴パラメータ

According to the present invention, the above object is solved by the following means.
(1) A method for diagnosing an abnormality in a part to be diagnosed in a rotary drive device.
1) At normal time (reference time) and at the time of diagnosis, the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part is processed by the statistical information filter, and the vibration characteristic data extracted at normal time and the diagnosis are performed, respectively. The first step to obtain time-extracted vibration characteristic data,
2) The normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data are normalized to obtain the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4, and the normal operation is performed. After performing the envelope processing on the extracted vibration characteristic data and the extracted vibration characteristic data at the time of diagnosis, each data is Fourier transformed to obtain the characteristic parameters related to the frequency domain for any one or more of the following equations 5 to 10. Second step and
3) Principal component analysis was performed on the characteristic parameters related to the time domain and the characteristic parameters related to the frequency domain obtained in the second step, and the overlapping ratio of the distribution areas at the time of normal diagnosis and the distribution area at the time of diagnosis in the obtained coordinate space of the principal components was used. This is a method for diagnosing an abnormality in a portion to be diagnosed in a rotary drive device, which comprises a third step of determining an abnormality in the drive unit in the rotary drive device.

a) Feature parameters related to the time domain

b) Feature parameters related to the time domain

c) Feature parameters related to the time domain

d) Feature parameters related to the time domain

e) Feature parameter representing the stability index of the low frequency band of the part to be diagnosed

f) Feature parameter representing the total power spectrum of the low frequency band of the diagnostic target

g) Feature parameter representing the path frequency component ratio of the part to be diagnosed

h) Feature parameter representing the damage index of the part to be diagnosed

i) Feature parameter representing the path frequency second harmonic rate of the part to be diagnosed

j) Characteristic parameter representing the path frequency third harmonic rate of the part to be diagnosed

With such a configuration, the vibration characteristic data obtained by the sensor installed at an appropriate position of the rotary drive device including the bearing can be processed by specific filtering, processing by specific parameters, and further, a method of principal component analysis. By processing by combining the above, it is possible to accurately extract an abnormal signal even if the S / N ratio is low as in the vibration characteristic data in a drive device including a bearing that rotates at a low speed. ..

(2) In the above item (1), at least one of the feature parameters related to the frequency domain for any one or more of the formulas 5 to 10 shall be selected from the feature parameters of the formulas 5 to 8.

With such a configuration, the invention of the above item (1) can be easily and efficiently carried out.

(3) In the above item (1) or (2), the principal component analysis shall be performed using a plurality of principal components whose total cumulative rate is equal to or higher than a predetermined value.

With such a configuration, the characteristics of the original data can be expressed well, and the accuracy of abnormality diagnosis can be improved.

(4) In any of the above items (1) to (3), the ratio at which the distribution area of the characteristic parameter at the time of diagnosis overlapped with the distribution area of the characteristic parameter at the time of diagnosis plotted in the coordinate space is predetermined. When it is less than or equal to the value, it is determined that the part to be diagnosed is abnormal.

With such a configuration, it is possible to easily and accurately determine the abnormality diagnosis.

(5) In any of the above items (1) to (4), it is assumed that the drive device is a chain transmission mechanism of a low-speed rotating machine.

With such a configuration, it is possible to easily perform an abnormality diagnosis in the chain transmission mechanism of a low-speed rotating machine, which has been difficult to diagnose in the past.

(6) In any of the above items (1) to (5), it is assumed that the diagnostic target portion in the rotary drive device is a bearing or a gear.

With such a configuration, according to the present invention, it is possible to easily and surely perform an abnormality diagnosis in a bearing or a gear, which has been difficult to diagnose in the past.

b)時間領域に係る特徴パラメータ

c)時間領域に係る特徴パラメータ

d)時間領域に係る特徴パラメータ

e) 診断対象部の低周波帯域の安定指数を表わす特徴パラメータ

f）診断対象部の低周波帯域のパワースペクトルの総和を表わす特徴パラメータ

g）診断対象部のパス周波数成分率を表わす特徴パラメータ

h）診断対象部の損傷指数を表わす特徴パラメータ

i）診断対象部のパス周波数２次高調波率を表わす特徴パラメータ

j）診断対象部のパス周波数３次高調波率を表わす特徴パラメータ

(7) An abnormality diagnosis device for a part to be diagnosed in a rotary drive device.
1) Normal extraction vibration to obtain normal extraction vibration characteristic data by processing the vibration characteristic data output from the sensor installed at the vibration measurement site of the diagnosis target part with a statistical information filter during normal operation (reference time). Characteristic data acquisition means and
2) As a means for acquiring vibration characteristic data extracted at the time of diagnosis, which obtains the vibration characteristic data extracted at the time of diagnosis by processing the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part with the statistical information filter at the time of diagnosis. ,
3) Acquisition of time domain feature parameters by normalizing the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, and obtaining the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4. Means and
4) After performing the envelope processing on the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, each data is Fourier transformed into the frequency domain for any one or more of the following equations 5 to 10. Frequency domain feature parameter acquisition means for obtaining such feature parameters,
5) Principal component analysis is performed on the characteristic parameters related to the obtained time domain and the characteristic parameters related to the frequency domain, and the rotation drive device is based on the overlapping ratio of the distribution areas at the time of normal diagnosis and the distribution area at the time of diagnosis in the obtained coordinate space of the main components. It is an abnormality diagnosis device of a diagnosis target part in a rotary drive device, which is provided with an abnormality determination means for determining an abnormality of the diagnosis target part in the above.

a) Feature parameters related to the time domain

b) Feature parameters related to the time domain

c) Feature parameters related to the time domain

d) Feature parameters related to the time domain

e) Feature parameter representing the stability index of the low frequency band of the part to be diagnosed

f) Feature parameter representing the total power spectrum of the low frequency band of the diagnostic target

g) Feature parameter representing the path frequency component ratio of the part to be diagnosed

h) Feature parameter representing the damage index of the part to be diagnosed

i) Feature parameter representing the path frequency second harmonic rate of the part to be diagnosed

j) Characteristic parameter representing the path frequency third harmonic rate of the part to be diagnosed

With such a configuration, the vibration characteristic data output from the sensor installed at an appropriate position of the rotary drive device is used to diagnose the bearing, gear, chain, etc. of the drive device that rotates at a low speed. An abnormality diagnostic device capable of reliably detecting an abnormality in a part is provided.

In the above item (7), it is assumed that the diagnostic target portion in the rotary drive device is a bearing or a gear.

With such a configuration, an abnormality diagnosis device capable of easily and surely performing an abnormality diagnosis in a bearing or a gear, which has been difficult to diagnose in the past, is provided.

The present invention uses vibration characteristic data output from sensors installed at appropriate locations in a rotary drive device to detect abnormalities in parts to be diagnosed, such as bearings, gears, and chains, especially in a drive device that rotates at a low speed. Can be reliably implemented.

It is a flow chart of the process of this invention. It is a figure which shows the low-speed rotating device and the vibration acceleration sensor as an example of the drive device including a bearing in this invention. It is a figure which shows an example of the outer ring scratch, the inner ring scratch, and the rolling body scratch of a bearing in this invention. As Comparative Example 1, it is a figure which shows the result when this invention was carried out without processing by a statistical information filter. As Example 1, it is a figure which shows the result when this invention was processed with a statistical information filter, and was carried out. As Comparative Example 2, it is a figure which shows the result when this invention was carried out without processing by a statistical information filter. As a second embodiment, it is a figure which shows the result when the present invention was processed by a statistical information filter, and was carried out. It is a figure which shows the material transport aircraft as an example of the drive device including a bearing in this invention, and the bearing to be diagnosed. It is a figure which shows the abnormality which occurred in the inner ring of the bearing to be diagnosed. It is a figure which shows the abnormality which occurred in the outer ring of the bearing to be diagnosed. It is a drawing which shows the principal component analysis result of Example 3. FIG. It is a drawing which shows the principal component analysis result of Example 4. FIG. It is a drawing which shows the principal component analysis result of Example 5. It is a drawing which shows the principal component analysis result of Example 6. It is a drawing which shows the principal component analysis result of Example 7. It is a drawing which shows the principal component analysis result of Example 8. It is a drawing which shows the principal component analysis result of Example 9. It is a drawing which shows the principal component analysis result of Example 10. It is a drawing which shows the principal component analysis result of Example 11. It is a drawing which shows the principal component analysis result of Example 12. It is a drawing which shows the principal component analysis result of Example 13. It is a drawing which shows the principal component analysis result of Example 14. It is a drawing which shows the principal component analysis result of Example 15. It is a drawing which shows the principal component analysis result of Example 16. It is a drawing which shows the principal component analysis result of Example 17. It is a drawing which shows the principal component analysis result of Example 18. It is a drawing which shows the principal component analysis result of Example 19. It is a drawing which shows the principal component analysis result of Example 20. It is a drawing which shows the principal component analysis result of Example 21. It is a drawing which shows the principal component analysis result of Example 22. It is a drawing which shows the principal component analysis result of Example 23. It is a drawing which shows the principal component analysis result of Example 24. It is a drawing which shows the principal component analysis result of Example 25. It is a drawing which shows the principal component analysis result of Example 26. It is a drawing which shows the principal component analysis result of Example 27. It is a drawing which shows the principal component analysis result of Example 28. It is a drawing which shows the principal component analysis result of Example 29. It is a drawing which shows the principal component analysis result of Example 30. It is a drawing which shows the principal component analysis result of Example 31. It is a drawing which shows the principal component analysis result of Example 32. It is a drawing which shows the principal component analysis result of Example 33. It is a drawing which shows the principal component analysis result of Example 34. It is a drawing which shows the principal component analysis result of Example 35. It is a drawing which shows the principal component analysis result of Example 36. It is a drawing which shows the principal component analysis result of Example 37. It is a drawing which shows the principal component analysis result of Example 38. It is a drawing which shows the principal component analysis result of Example 39. It is a drawing which shows the principal component analysis result of Example 40. It is a drawing which shows the principal component analysis result of Example 41. It is a drawing which shows the principal component analysis result of Example 42. It is a drawing which shows the principal component analysis result of Example 43. It is a drawing which shows the principal component analysis result of Example 44. It is a drawing which shows the principal component analysis result of Example 45. It is a drawing which shows the principal component analysis result of Example 46. It is a drawing which shows the principal component analysis result of Example 47. It is a drawing which shows the principal component analysis result of Example 48. It is a figure which shows the principal component analysis result of the comparative example 3. FIG.

Hereinafter, embodiments of the present invention will be described in detail. FIG. 1 is a flow chart of the process of the present invention.
In the present invention, the position where the sensor is attached to the drive device including the bearing, that is, the vibration measuring portion is not particularly limited as long as it is a position where the vibration of the bearing can be detected. In general, a fixing member that directly or indirectly contacts a bearing, specifically, a casing of a drive device, an installation base, or the like is suitable.
In the present invention, a known sensor is used without particular limitation as the sensor used for detecting the vibration characteristic data.

(1) First, the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target portion is processed by the statistical information filter at the time of normal (reference time) and at the time of diagnosis, and the vibration characteristics extracted at normal time are respectively. The first step of obtaining the data and the vibration characteristic data extracted at the time of diagnosis will be described.

In the usual case, the diagnostic method of the present invention first acquires normal vibration characteristic data. Here, the normal state (reference time) means a state in which no abnormality has occurred as a result of performing a detailed diagnosis of the equipment to be diagnosed. In order to improve accuracy, it is preferable to acquire the diagnostic signal for a sufficient time length (for example, 5 rotations).

Next, the acquired normal vibration characteristic data is processed by a statistical information filter in order to remove noise.
The processing by the statistical information filter is performed after converting the vibration characteristic data, which is a time waveform output from the sensor, into a spectral waveform by Fourier transform, preferably fast Fourier transform. Prior to the Fourier transform, it is preferable to normalize the data and divide the data, if necessary, in order to efficiently perform the processing.

Normalization of the above data is carried out by the following equation 11.

In the present invention, the spectrum data processed by the above statistical information filter is processed by a high-pass filter if necessary, and then converted into a time waveform by an inverse Fourier transform, preferably an inverse fast Fourier transform, to obtain normal-time extracted vibration characteristic data. ..

In the diagnostic method of the present invention, usually, after acquiring the normal vibration characteristic data, the diagnostic vibration characteristic data to be diagnosed is acquired.
Similar to the normal vibration characteristic data, the above diagnostic vibration characteristic data is acquired so as to measure a diagnostic signal for a sufficient time length (for example, for 5 rotations) in order to perform statistical processing and improve accuracy. It is preferable to do so. Further, in the acquisition of data, the data may be continuously collected and the target data may be extracted from the data, or the required amount of data may be acquired only at the time of diagnosis.

Next, the acquired vibration characteristic data at diagnosis is processed by a statistical information filter in the same manner as the vibration characteristic data at normal time.
Similarly, the spectrum data processed by the above statistical information filter is processed by a high-pass filter if necessary, and then converted into a time waveform by an inverse Fourier transform, preferably an inverse fast Fourier transform, and the vibration characteristic data extracted at the time of diagnosis is obtained. obtain.
It should be noted that the noise removal by the processing of the statistical information filter is not sufficient, and the obtained extracted vibration characteristic data is processed so that the abnormal data can be easily detected by the following processing.
As described above, normally, after acquiring the normal vibration characteristic data, the diagnostic vibration characteristic data is acquired, but the present invention is not limited to this, and after the diagnostic vibration characteristic data is acquired, the normal vibration characteristic data is acquired. The present invention may be carried out by obtaining the data.

(2) Next, the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data are normalized to obtain the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4. After performing the envelope processing on the vibration characteristic data extracted at the time of normal and the vibration characteristic data extracted at the time of diagnosis, each data is subjected to Fourier transform to relate to the frequency domain for any one or more of the following equations 5 to 10. The second step of obtaining the feature parameters will be described.

Normalization is performed by the following equation 11.

b)時間領域に係る特徴パラメータ

c)時間領域に係る特徴パラメータ

d)時間領域に係る特徴パラメータ

Next, the feature parameters related to the time domain are obtained by the following equations 1 to 4.
a) Feature parameters related to the time domain

b) Feature parameters related to the time domain

c) Feature parameters related to the time domain

d) Feature parameters related to the time domain

In order to obtain the envelope processing for the vibration characteristic data extracted during normal operation and the vibration characteristic data extracted during diagnosis, the measured time series signal is subjected to a high-pass filter, and then the envelope of the waveform is obtained.
Also, each data is converted into frequency data. That is, after the envelope is obtained, the spectrum of the envelope is obtained by Fourier transform.
Further, the feature parameters related to the frequency domain for any one or more of the following equations 5 to 10 are obtained.

f）診断対象部の低周波帯域のパワースペクトルの総和を表わす特徴パラメータ

g）診断対象部のパス周波数成分率を表わす特徴パラメータ

h）診断対象部の損傷指数表わす特徴パラメータ

i）診断対象部のパス周波数２次高調波率を表わす特徴パラメータ

j）診断対象部のパス周波数３次高調波率を表わす特徴パラメータ

e) Feature parameter representing the stability index of the low frequency band of the part to be diagnosed

f) Feature parameter representing the total power spectrum of the low frequency band of the diagnostic target

g) Feature parameter representing the path frequency component ratio of the part to be diagnosed

h) Characteristic parameters representing the damage index of the part to be diagnosed

i) Feature parameter representing the path frequency second harmonic rate of the part to be diagnosed

j) Characteristic parameter representing the path frequency third harmonic rate of the part to be diagnosed

(3) Next, principal component analysis is performed on the characteristic parameters related to the time domain and the characteristic parameters related to the frequency domain obtained in the second step, and the distribution areas at the time of normal and diagnosis in the obtained coordinate space of the principal components are performed. The third step of determining the abnormality of the drive unit in the rotary drive device based on the overlap ratio of the above will be described.

First, principal component analysis is performed on the characteristic parameters in the time domain and the characteristic parameters in the frequency domain.

であれば、「正常」と判定し、

The obtained principal component is judged to be abnormal in the drive device based on the overlapping ratio of the distribution areas on the coordinate axes.

If so, it is judged as "normal" and

The present invention also provides an abnormality diagnosis device used in the above-mentioned abnormality diagnosis method. That is, it is an abnormality diagnosis device of the diagnosis target part in the rotation drive device.
1) Normal extraction vibration to obtain normal extraction vibration characteristic data by processing the vibration characteristic data output from the sensor installed at the vibration measurement site of the diagnosis target part with a statistical information filter during normal operation (reference time). Characteristic data acquisition means and
2) As a means for acquiring vibration characteristic data extracted at the time of diagnosis, which obtains the vibration characteristic data extracted at the time of diagnosis by processing the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part with the statistical information filter at the time of diagnosis. ,
3) Acquisition of time domain feature parameters by normalizing the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, and obtaining the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4. Means and
4) After performing the envelope processing on the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, each data is Fourier transformed into the frequency domain for any one or more of the following equations 5 to 10. Frequency domain feature parameter acquisition means for obtaining such feature parameters,
5) Principal component analysis is performed on the characteristic parameters related to the obtained time domain and the characteristic parameters related to the frequency domain, and the rotation drive device is based on the overlapping ratio of the distribution areas at the time of normal diagnosis and the distribution area at the time of diagnosis in the obtained coordinate space of the main components. Provided is an abnormality diagnosis device for a diagnosis target part in a rotation drive device, which comprises an abnormality determination means for determining an abnormality of the diagnosis target part in the above.
The meanings of Equations 1 to 4 for the feature parameters related to the time domain, Equations 5 to 10 for the feature parameters related to the frequency domain, and the symbols are the same as those in the above-mentioned abnormality diagnosis method.

The above-mentioned abnormality diagnosis method and abnormality diagnosis device can be similarly applied to the state diagnosis of these parts by defining the characteristic parameters for diagnosis of other parts (gears, slide bearings, etc.).

(Examples 1 and 2) (Comparative Examples 1 and 2)
In order to measure the vibration of the bearing holder of the low-speed rotating machine simulator shown in Fig. 2, the acceleration signal of the vibration was measured with a sensor attached at an appropriate position.
The sampling frequency was 100 kHz, and the sampling time was 200 sec when measuring 10 to 40 rpm and 60 sec when measuring 50 to 100 rpm.
In addition, assuming an early abnormality in the field, we prepared three types of bearings (outer ring scratch, inner ring scratch, rolling body scratch) with spot scratches of width 5.0 mm x depth 0.5 mm as shown in Fig. 3. ..
The signal obtained by the experiment is processed by the method of the present invention, and the final result of the principal component analysis is shown. In order to show the effectiveness of this method, the results of 10 rpm and 70 rpm are posted in a form that compares before and after the statistical information filtering process.
First, for 10 rpm, as shown in FIG. 4 (Comparative Example 1: Before statistical information filter processing) and FIG. 5 (Example 1: After statistical information filter processing), the first principal component is in abscissa and the second main component. The components are in ordinates and are shown in the coordinate space. The lower figure in each figure is an enlarged view of the vicinity of the origin of the coordinates in the upper figure. The same applies to the following examples and comparative examples. The main component values in the normal state (indicated by ○) are concentrated near the origin, and the main component values in the normal state and the main component values in the abnormal state (□ mark for outer ring scratches, △ mark for inner ring scratches). It is clear that the distance from the rolling body injury (indicated by a cross) is significantly improved after the statistical information filter processing than before the processing.
Next, at 70 rpm, as shown in Fig. 6 (Comparative example 2: before statistical information filter processing), even in the results before the statistical information filter, there is a tendency for each group to solidify, but it is in the normal state (marked with a circle). The group of abnormal conditions (□ mark, △ mark, × mark) overlaps the group, and it is not possible to make an accurate diagnosis. However, as shown in FIG. 7 (Example 2: after statistical information filtering), it is clear that the identification can be made clearly after the statistical information filtering.
Next, an example of verification using on-site equipment is shown. In order to confirm the effectiveness of the method devised in the present invention, verification with on-site equipment was also performed. Here, an example of verification is shown.
FIG. 8 shows the equipment to be verified, and Table 1 shows the bearing specifications. Since there is a main chain wheel near the bearing to be diagnosed and the roller chain (with bucket) is driven, it is accompanied by shocking vibration even in the normal state.
As shown in FIGS. 9 and 10, the bearing to be diagnosed had flaking and scratches on the inner ring and the outer ring. The vibration acceleration signal is measured before and after the bearing is replaced, and the result of diagnosis using the method devised in the present invention is shown in FIG. 11 (Example 3). The main component value of the abnormal state before the bearing replacement is far from the normal state after the replacement, and it is clear that the abnormal state can be clearly detected and diagnosed.

(Examples 4 to 48)
Similar to Examples 1 and 2, the vibration characteristic signal was measured by a sensor installed in the bearing holder, which is the diagnostic target portion of the low-speed rotating machine simulator shown in FIG.
Similar to Examples 1 and 2, there are three types of spot scratches (outer ring scratch, inner ring) with a width of 5.0 mm and a depth of 0.5 mm, as shown in FIG. 3, for measuring the vibration characteristic signal at the time of diagnosis. Scratches, rolling body scratches) bearings were used.
The sampling frequency was the same as in Examples 1 and 2, and the rotation speed was 10 rpm.
As shown in Tables 2-1 to 2-3, various combinations of the formula for obtaining the feature parameter related to the time domain and the formula for obtaining the feature parameter related to the frequency domain were processed by the method of the present invention. In addition, the numbers of the drawings in which the obtained principal component analysis results are plotted in the coordinate space are shown in the table.
From these drawings of FIGS. 12 to 56, the effectiveness of the present invention is clear in any combination.

(Comparative Example 3)
The present invention is the same as in Examples 4 to 48, except that only equations 5 and 7 are used among the equations for obtaining the characteristic parameters related to the frequency domain without using the equation for obtaining the characteristic parameters related to the time domain. The result of processing by the method of FIG. 57 is shown in FIG. The group of abnormal states (marked with □, △, and ×) overlaps the group of normal states (marked with ○), and it is not possible to diagnose whether the state is normal or abnormal. There is.

Claims (8)

This is a method for diagnosing abnormalities in the part to be diagnosed in the rotary drive device.
1) At normal time (reference time) and at the time of diagnosis, the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part is processed by a noise removal filter that removes noise, and the vibration extracted at normal time respectively. The first step to obtain characteristic data and vibration characteristic data extracted at the time of diagnosis,
2) The normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data are normalized to obtain the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4, and the normal operation is performed. After performing the envelope processing on the extracted vibration characteristic data and the extracted vibration characteristic data at the time of diagnosis, each data is Fourier transformed to obtain the characteristic parameters related to the frequency domain for any one or more of the following equations 5 to 10. Second step and
3) Principal component analysis was performed on the characteristic parameters related to the time domain and the characteristic parameters related to the frequency domain obtained in the second step, and the overlapping ratio of the distribution areas at the time of normal diagnosis and the distribution area at the time of diagnosis in the obtained coordinate space of the principal components was used. A method for diagnosing an abnormality in a portion to be diagnosed in a rotary drive device, which comprises a third step of determining an abnormality in the drive unit in the rotary drive device.

a) Feature parameters related to the time domain

b) Feature parameters related to the time domain

c) Feature parameters related to the time domain

d) Feature parameters related to the time domain

e) Feature parameter representing the stability index of the low frequency band of the part to be diagnosed

f) Feature parameter representing the total power spectrum of the low frequency band of the diagnosis target part

g) Feature parameter representing the path frequency component ratio of the part to be diagnosed

h) Feature parameter representing the damage index of the part to be diagnosed

i) Feature parameter representing the path frequency second harmonic rate of the part to be diagnosed

j) Characteristic parameter representing the path frequency third harmonic rate of the part to be diagnosed

The diagnostic target portion in the rotary drive device according to claim 1, wherein at least one of the characteristic parameters related to the frequency domain for any one or more of the equations 5 to 10 is selected from the characteristic parameters of the equations 5 to 8. Abnormality diagnosis method. When the number of principal components is m, the total from the standard deviation of the first principal component to the standard deviation of the i-principal component, which is the total value from the standard deviation of the first principal component to the standard deviation of the m-th principal component. The rotational drive according to claim 1 or 2 , wherein the principal component analysis is performed using the first principal component to the i principal component such that the cumulative rate, which is a ratio to the value, is equal to or more than a predetermined predetermined value. A method for diagnosing abnormalities in the part to be diagnosed in the device. Claim that the diagnosis target part is judged to be abnormal when the distribution area of the feature parameter at the time of diagnosis plotted in the coordinate space is equal to or less than a predetermined value in the distribution area of the feature parameter at the time of normal diagnosis. Item 3. The method for diagnosing an abnormality in a portion to be diagnosed in the rotary drive device according to any one of Items 1 to 3. The method for diagnosing an abnormality in a portion to be diagnosed in a rotary drive device according to any one of claims 1 to 4, wherein the drive device is a chain transmission mechanism of a low-speed rotary machine. The method for diagnosing an abnormality in a portion to be diagnosed in a rotary drive device according to any one of claims 1 to 5, wherein the diagnostic target portion in the rotary drive device is a bearing or a gear. It is an abnormality diagnosis device of the diagnosis target part in the rotation drive device.
1) In the normal state (reference time), the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part is processed by a noise removal filter for removing noise to obtain the vibration characteristic data extracted in the normal state. Normal extraction vibration characteristic data acquisition means and
2) As a means for acquiring vibration characteristic data extracted at the time of diagnosis, which obtains the vibration characteristic data extracted at the time of diagnosis by processing the vibration characteristic data output from the sensor installed at the vibration measurement part of the diagnosis target part with the statistical information filter at the time of diagnosis. ,
3) Acquisition of time domain feature parameters by normalizing the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, and obtaining the characteristic parameters related to the time domain for any one or more of the following equations 1 to 4. Means and
4) After performing the envelope processing on the normal extraction vibration characteristic data and the diagnosis extraction vibration characteristic data, each data is Fourier transformed into the frequency domain for any one or more of the following equations 5 to 10. Frequency domain feature parameter acquisition means for obtaining such feature parameters,
5) Principal component analysis is performed on the characteristic parameters related to the obtained time domain and the characteristic parameters related to the frequency domain, and the rotation drive device is based on the overlapping ratio of the distribution areas at the time of normal diagnosis and the distribution area at the time of diagnosis in the obtained coordinate space of the main components. An abnormality diagnosis device for a diagnosis target part in a rotation drive device, which comprises an abnormality determination means for determining an abnormality of the diagnosis target part in the above.

a) Feature parameters related to the time domain

b) Feature parameters related to the time domain

c) Feature parameters related to the time domain

d) Feature parameters related to the time domain

e) Feature parameter representing the stability index of the low frequency band of the part to be diagnosed

f) Feature parameter representing the total power spectrum of the low frequency band of the diagnosis target part

g) Feature parameter representing the path frequency component ratio of the part to be diagnosed

h) Feature parameter representing the damage index of the part to be diagnosed

i) Feature parameter representing the path frequency second harmonic rate of the part to be diagnosed

j) Characteristic parameter representing the path frequency third harmonic rate of the part to be diagnosed

The abnormality diagnosis device for a diagnosis target portion in a rotation drive device according to claim 7, wherein the diagnosis target portion in the rotation drive device is a bearing or a gear.

Images