JP3214233B2 - Rotating machine vibration diagnostic device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for diagnosing vibration of a rotating machine for inspecting the vibration of the rotating machine, diagnosing and detecting abnormalities, and determining the prediction thereof.

[0002]

2. Description of the Related Art Many abnormalities during the rotation of a general rotating device often appear as changes in shaft vibration or noise, which may cause a serious accident in a large rotating device. Recently, there has been proposed a rotating machine vibration diagnosis apparatus that detects vibration abnormalities or abnormal signs and diagnoses vibrations online. FIG. 8 shows a configuration diagram of a conventional rotating machine vibration diagnosis device 1. In the figure, a bearing 2 supporting a rotor 2 of a rotating machine has a rotor 2
A vibrometer 4 for detecting the vibration of the rotor 2 is provided, and a tachometer 5 for detecting the rotation speed of the rotor 2 is provided at one end of the rotor 2.

Further, the rotating machine is provided with a bearing thermometer 6 and various sensors 7 for monitoring the operating state of the rotating machine, such as input voltage and current. Each of these instruments 4, 5,
Signals from the sensors 6 and 7 are sampled and input by the signal sampling means 8 of the rotating machine vibration diagnostic device 1. Here, the signal used for diagnosis and analysis of rotation abnormality is only the vibration signal of the bearing, and this signal is usually analyzed in various ways,
Diagnosis is based on the results. For example, as shown in FIG. 8, a vibration signal S is obtained from the vibration meter 4 and this signal is
FT (fast Fourier transform) is performed. As a result, it is confirmed that in the component of the vibration signal S, there is a rotational component synchronized with the rated rotational speed of the rotor and its harmonics.

The data processing means 9 performs calculations for detecting and diagnosing abnormalities, and further performs FFT to obtain a frequency spectrum, a vibration phase, and a rate of change thereof.
The abnormality detecting means 10 sets a threshold value for these calculated values, detects an abnormal sign of vibration, and activates the diagnosing means 11. After the abnormality detection, the diagnosis unit 11
Abnormality estimation is performed using the above-described various calculated values, correlation with the operation state, and the like.

[0005] The display control means 12 outputs the result of the data processing.
The notification of the abnormality detection and the result of the diagnosis are displayed on the display device 13. Further, the history data is recorded in the history data file 14B by the data recording means 14A for recording the data output by the data processing means 9, and the history data file 14 is recorded.
A history data storage system 14 having data reproducing means 14C for reproducing data from B is added.

Further, in order to determine the cause of the occurrence of the abnormal phenomenon from the frequency spectrum and the vibration phase obtained by the data processing means 9 and the rate of change thereof, it is necessary to analyze a huge amount of data by an expert. History data storage system 1
No. 4 requires a large-capacity storage device for storing an enormous amount of data, and cannot be handled by small-scale hardware. Further, long-term data storage requires data management, such as periodically transferring data to another storage device.

[0007]

The conventional rotating machine vibration diagnostic apparatus 1 uses only the bearing vibration data for these analyses. Insufficiently, bearing noise, electromagnetic noise, and the like, which change due to the structure of the main body and variable speed operation, may be the main cause, and it may be necessary to analyze the vibration data on which these noises are based.

[0008] However, there is a method of detecting the operating condition by sound as a means for judging the driving condition. However, since there is also a random noise around the device, to increase the detection level, external vibration and noise are cut off and sound is absorbed inside. A processed soundproofing and sound absorbing box or room is required, the equipment becomes large-scale, the equipment cost is high, and there is a limit in increasing the analysis level even when relying on primitive hearing.

In addition, stable measurement is difficult in ordinary vibration measurement, and a large amount of random noise components may be contained. Noise removal is important for diagnosis and prediction of abnormality. A dedicated filter and microcomputer, and a large-scale and expensive dedicated vibration analysis device of the mini-con class are required.

[0010] In addition, the need for long-term storage of signal data, high-speed waveform processing, and analysis by specialists increases for on-line diagnosis, and the analysis takes time, and real-time diagnosis during operation is performed. There were problems such as not being able to do. In addition, inexpensive and small-scale hardware cannot cope with processing of an enormous amount of data.

A first object of the present invention is to provide an inexpensive rotating machine vibration diagnosis apparatus capable of detecting an abnormal vibration of a rotating machine online. To provide.

It is a second object of the present invention to provide a rotating machine vibration diagnostic apparatus capable of predicting a time required to reach an abnormal area and a dangerous area of the rotating machine vibration. A fourth object of the present invention is to provide a rotating machine vibration diagnostic apparatus capable of predicting a noise frequency component value corresponding to the frequency from a vibration frequency component value obtained and detecting an abnormal noise on-line. A fifth object of the present invention is to provide a rotating machine vibration diagnosis device capable of quickly diagnosing the cause of abnormal vibration of a machine, and a fifth object is to predict a noise frequency component value corresponding to a frequency from a measured vibration frequency component value. It is another object of the present invention to provide a rotating machine vibration diagnosis device capable of quickly diagnosing the cause of abnormal noise.

[0013]

SUMMARY OF THE INVENTION A rotating machine vibration diagnostic apparatus according to the present invention comprises: a rotating speed detecting means for detecting a rotating speed of a rotating machine; a vibration detecting means for detecting vibration of the rotating machine; A signal sampling means for sampling a signal output by the means for a predetermined period to output a frequency spectrum; a noise removing means for outputting a frequency spectrum obtained by removing noise from the frequency spectrum output by the signal sampling means; A deterioration index processing means for calculating a vibration deterioration index which is a ratio of a rotation synchronization frequency component value synchronized with the rotation speed output by the rotation speed detection means and an integral multiple frequency component value of the output frequency spectrum; Abnormality diagnosis processing means for outputting a signal when the deterioration index reaches a predetermined value.

A rotating machine vibration diagnostic apparatus according to the next invention is a deterioration index storing means for storing a history value of the vibration deterioration index, and calculates a vibration deterioration index at an arbitrary time from the history value of the vibration deterioration index. This is provided with deterioration prediction means.

Further, a rotating machine vibration diagnostic apparatus according to the next invention comprises a deterioration index storing means for storing a vibration deterioration index of a characteristic parameter for each vibration cause of the rotating machine, and a frequency spectrum outputted by the noise removing means. Then, the tendency of the vibration deterioration index calculated from the above is compared with the vibration deterioration index of the characteristic parameter for each vibration cause to diagnose the vibration cause.

Also, a rotating machine vibration diagnosis apparatus according to the next invention comprises a rotating speed detecting means for detecting the rotating speed of the rotating machine, a vibration detecting means for detecting the vibration of the rotating machine, and an output of the vibration detecting means. Signal sampling means for sampling a signal to be output for a predetermined period to output a frequency spectrum, noise removing means for outputting a frequency spectrum obtained by removing noise from the frequency spectrum output by the signal sampling means, and a specific noise included in the rotating machine. It has related data of a frequency component value and a vibration frequency component value corresponding to this frequency, and predicts a noise frequency component value from the vibration frequency component value output by the noise removing means by using the related data. Noise prediction means; and abnormality diagnosis processing means for outputting a signal when the noise frequency component value reaches a predetermined value.

The rotating machine vibration diagnostic apparatus according to the next invention comprises a noise frequency component value storage means for storing a history value of a predetermined noise frequency component value output by the noise prediction means, and the predetermined noise frequency component value. And a deterioration prediction means for calculating the predetermined noise frequency component value at an arbitrary time from the history value of the above.

Further, in the rotating machine vibration diagnostic apparatus according to the next invention, the noise removing means outputs a frequency spectrum obtained by averaging repeated frequency components in a frequency spectrum outputted by the sampling means. .

[0019]

According to the rotary machine vibration diagnosis apparatus of the present invention, a ratio of a rotation synchronization frequency component value synchronized with the rotation speed output by the rotation speed detection unit to an integral multiple frequency component value of the frequency spectrum output by the noise removal unit is provided. And a failure diagnosis processing means for outputting a signal when the vibration deterioration index reaches a predetermined value, so that an integer multiple determined for each abnormal phenomenon unique to the rotating machine is provided. It operates when the vibration deterioration index of the frequency component reaches a predetermined value.

Further, a rotating machine vibration diagnostic apparatus according to the next invention has a deterioration index storing means for storing a history value of the vibration deterioration index, and a deterioration prediction for calculating a vibration deterioration index at an arbitrary time from the history value of the vibration deterioration index. Means, the vibration deterioration index at any time is predicted.

Further, the rotating machine vibration diagnostic apparatus according to the next invention comprises a deterioration index storing means for storing a vibration deterioration index of a characteristic parameter for each vibration cause of the rotating machine, and uses a frequency spectrum output from the noise removing means. The calculated tendency of the vibration deterioration index is compared with the vibration deterioration index of the characteristic parameter for each vibration cause to diagnose the vibration cause.

Further, a rotating machine vibration diagnostic apparatus according to the next invention has data relating a specific noise frequency component value of the rotating machine and a vibration frequency component value corresponding to this frequency, and the noise removing means is provided. A noise prediction means for predicting a noise frequency component value from the output vibration frequency component value using related data and an abnormality diagnosis processing means for outputting a signal when the noise frequency component value reaches a predetermined value are provided. In addition, even at a site with high background noise, a noise level abnormality is detected without providing a sound insulation device.

Further, a rotating machine vibration diagnosis apparatus according to the next invention comprises a noise frequency component value storage means for storing a history value of a predetermined noise frequency component value output by the noise prediction means, and a history of the predetermined noise frequency component value. Since there is provided deterioration prediction means for calculating a predetermined noise frequency component value at an arbitrary time from the value, the predetermined noise frequency component value at an arbitrary time is predicted.

Further, in the rotating machine vibration diagnosis apparatus according to the next invention, the noise elimination means outputs a frequency spectrum obtained by averaging repeated frequency components in the frequency spectrum output from the sampling means. Is removed.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows a configuration diagram of a rotating machine vibration diagnostic apparatus used for an electric motor according to one embodiment of the present invention. In the figure, the same reference numerals as those in FIG. In the figure, reference numeral 4A denotes a vibrometer provided at the center of the outer frame of the electric motor, which detects vibration of the electric motor body and vibration caused by magnetic excitation force generated by interference between a stator (not shown) and a magnetic field of the rotor 2. I do. Reference numeral 8 denotes signal sampling means for sampling the vibration signals of the vibrometers 4 and 4A in synchronization with the signal of the tachometer 5 for detecting the rotation speed of the rotor 2, and 14 denotes a frequency component value of the sampled vibration signal which is repeated. Is a noise removing means for outputting a frequency spectrum obtained by adding and averaging (cross averaging method). still,
The effect of noise elimination increases with the number of times of averaging, but the effect is not significant when the number of averaging is about 5 to 7 or more. In this embodiment, noise is removed by averaging seven times.

Numeral 15 stores data relating to noise and vibration for each frequency generated, such as bearing noise and magnetic noise, which are unique to the motor, and corresponds to the noise by multiplying the predicted metric value R for each measured vibration frequency component. A noise prediction unit 17 for predicting a frequency component value includes a rotation synchronization frequency component N synchronized with the rotation speed of the rotor 2 and 1 / 2N, 2N, 3N, 4N, among the frequency spectrum output from the noise removal unit 14. An integer multiple frequency component such as 5N and a phase are analyzed, and the rotation synchronization frequency component value P (fN1) and 1 / 2N, 2N, 3N, 4N
This is a deterioration index processing means for calculating a vibration deterioration index (pn) by a ratio with respect to an integer multiple frequency component value P (fNn) such as N, 5N or the like according to the following equation. D (pn) = P (fNn) / P (fN1) (1)

Numeral 18 compares each characteristic parameter (pn) of some vibration causes with a predetermined threshold value of a predetermined vibration deterioration index (pn), thereby obtaining characteristic parameters.
Diagnosis of vibration degradation index (pn)
Diagnostic processing means 19 which outputs a signal when the threshold value reaches a threshold value, is calculated by the deterioration index recording means 20, and calculates the vibration deterioration index (pn) for each time history recorded in the deterioration index data file 21. This is a deterioration predicting means for obtaining a regression coefficient by calculating the vibration regression coefficient and predicting a vibration deterioration index (pn) or a vibration level at an arbitrary time thereafter. Reference numeral 16 denotes an abnormality diagnosis prediction system including a deterioration index processing unit 17, an abnormality diagnosis processing unit 18, a deterioration prediction unit 19, a deterioration index recording unit 20, and a deterioration index data file 21, and 1A denotes the rotating machine diagnosis device. On the display device 13, the processing results of the deterioration index processing means 17, the abnormality diagnosis processing means 18, and the deterioration prediction means 19 can be displayed through the display means 12.

Next, the operation will be described with reference to the drawings. FIG. 2 is an explanatory diagram showing the relationship between the noise level predicted by the noise prediction unit 15 and the measured vibration level. In the figure, the horizontal axis represents the vibration level (dB) and the vertical axis represents the noise level (dB). The vibration level (dB) and the noise level (dB) of the rotating machine are shown.
Are measured simultaneously, and Vib (fm1) and SPLc (fm
1) and a function fm1 (Rfm1) is obtained from two points Vib (fm2) and SPLc (fm2).
5 is stored in a storage device (not shown). Similarly, the function fm2 (Rfm2) is obtained for the vibration level (dB) and the noise level (dB) of the other frequency components, and stored in the storage device of the noise prediction unit 15. In this way, the relationship between the vibration level (dB) and the noise level (dB) for the items for which the abnormality of the rotating machine must be determined by noise is stored in advance, and the noise abnormality is diagnosed using the related data. For example, when the vibration level (dB) of a certain frequency component is V
By arranging that the noise level (dB) becomes an abnormal value SPLc (fm2) when the noise level reaches ib (fm2), it is possible to detect noise abnormality. Therefore, when measuring the noise level (dB), a device for blocking background noise is required. Even without such a device, the noise level (dB) can be calculated by the noise prediction means 15 from the measured vibration level (dB). dB) can be predicted.

[0029]

FIG. 3 shows a processing flow of the abnormality diagnosis processing means 18.
FIG. In the figure, the abnormality diagnosis processing means 18 determines whether the vibration over-all level has increased (step S1), the rotational synchronization component level has increased (step S2), or the harmonic component has increased. (Step S3) It is determined whether the vibration deterioration index (pn) is increasing (Step S4), and if any of them continues to increase, or if any of them exceeds the threshold value And outputs a signal corresponding to the content of the abnormality, and the display means 12 displays the content of the abnormality on a display device. Further, the abnormality diagnosis processing means 18 calculates the frequency component 1 synchronized with the rotation speed for each vibration cause shown in Table 1 stored in the deterioration index data file 21 as the deterioration index storage means.
N and a vibration degradation index (pn) corresponding to the vibration level rate at which each of the frequency components 2N, 3N, 5N, and 1N or less, and a vibration degradation index (pn) processed by the degradation index processing means 17 By comparing the tendency with the tendency, the most similar vibration cause is selected, the display content is instructed on the display means 12, and the vibration cause and the vibration deterioration index (pn) are displayed on the display device 13, so that the abnormality diagnosis of the rotating machine can be performed. Do. As described above, since the abnormality diagnosis processing means 18 compares the vibration deterioration index (pn) corresponding to the characteristic parameter (pn) for each vibration cause, the characteristic parameter (pn) is stored in the deterioration index data file 21. ) May be stored, the capacity of the deterioration index data file 21 can be reduced, and the number of data to be compared can be reduced, so that quick processing can be performed. effective.

FIG. 4 is a flow chart showing the processing flow of the deterioration predicting means 19, and FIG. 5 shows the relationship between the temporal change of the vibration deterioration index (pn) and the threshold value of the vibration deterioration index (pn). FIG. In the figure, the deterioration prediction means 19
When the characteristic parameter (pn) of the vibration deterioration index (pn) to be predicted is specified (step S5), the sampling period t1 to t2 shown in FIG. 5 of the vibration deterioration index (pn) is selected (step S5). S6), the vibration deterioration index (p collected during this period and stored in the deterioration index data file 21)
n), the regression coefficient is calculated from the time history value of the vibration deterioration index (pn), and the vibration deterioration index (pn) at an arbitrary time after time t2 is calculated (step S).
7). Next, the deterioration predicting means 19 reaches predetermined threshold values of the safety zone, the abnormal zone, and the danger zone of the vibration deterioration index (pn) stored in the deterioration index data file 21 shown in FIG. The times ta and ts are calculated, the time to reach the abnormal area and the dangerous area is predicted (step S8), the display contents are instructed on the display means 12, and the contents of FIG. Although FIG. 5 shows the change of the vibration deterioration index (pn), the change may be indicated by the vibration level. As described above, the history value of the vibration deterioration index (pn) stored in the deterioration index data file 21 may be the vibration deterioration index (pn) corresponding to the characteristic parameter (pn) for each vibration cause. It is not necessary to store the history values of all the spectra, the capacity of the deterioration index data file 21 can be reduced, and the number of data to be compared can be reduced, so that the processing can be performed quickly.

FIGS. 6 and 7 show the quality judgment of the rotating machine in the initial state based on the vibration deterioration index (pn) of the characteristic parameter (pn) due to the distortion or loss of the shaft, and the time history value of the vibration deterioration index (pn). FIG. 9 is an explanatory diagram showing a case where abnormality prediction is performed from FIG. In the drawing, it is shown that the quality of the rotating machine is determined based on the relationship with the threshold for quality determination indicated by the broken line at the initial time t1. By lowering the threshold value of the abnormality diagnosis processing means 18 in this way, it can be used to determine the quality of the product shipping inspection. FIGS. 6 and 7 show the case where the characteristic parameter (pn) is designated and the judgment is made based on the vibration deterioration index (pn) of the characteristic parameter (pn). If the abnormality diagnosis apparatus 1A detects a vibration abnormality at the stage of shipping inspection without instructing (pn), the abnormality diagnosis processing means 18 is further operated, and the vibration index stored in the deterioration index data file 21 is stored. Degradation index (p
The cause of vibration can be diagnosed by comparing the tendency with n). FIG. 7 shows a case where the time history value of the vibration deterioration index (pn) has an inflection point at a certain point in time. However, by changing the sampling period, the regression coefficient is recalculated to obtain the prediction accuracy. Can be improved.

As described above, the vibration deterioration index (pn) is processed by the abnormality diagnosis processing means 18 and the deterioration prediction means 19, and the result is displayed on the display device 13 to diagnose the cause of vibration of the rotating machine. Quality improvement and vibration abnormal region,
Since it is possible to predict the time required to reach the danger zone, it is possible to quickly determine an appropriate action to avoid a serious accident and a timing of the action.

In this embodiment, the vibration deterioration index (p
Although the example of performing abnormality diagnosis and prediction using n) has been described,
The deterioration index data file 21 stores the abnormal region, the threshold value of the dangerous region, and the threshold value of the pass / fail judgment of the characteristic noise frequency level of the rotating machine. Are stored in the deterioration index data file 21 which is also a noise frequency component value storage device, and the stored data are stored.
The abnormality diagnosis unit 18 and the deterioration prediction unit 18 compare and process the noise level and the noise level predicted by the noise prediction unit 15 to diagnose the cause of the noise and predict the time when the noise level reaches the abnormal region and the dangerous region. Therefore, it is necessary to make a determination based on a noise level different from the sign of vibration of the rotor. For example, it is possible to quickly perform determination, diagnosis, and abnormality prediction of bearing noise and magnetic noise of the electric motor.

[0035]

As described above, according to the present invention, the rotating machine vibration diagnosing device is provided with a rotational synchronization frequency component synchronized with the rotational speed output from the rotational speed detecting means in the frequency spectrum output from the noise removing means. When the apparatus is provided with a deterioration index processing means for calculating a vibration deterioration index which is a ratio between the vibration deterioration index and the integral multiple frequency component value, and an abnormality diagnosis processing means for outputting a signal when the vibration deterioration index reaches a predetermined value, the rotating machine has When the vibration deterioration index of the integral multiple frequency component determined for each abnormal phenomenon becomes a predetermined value, the apparatus operates at a low cost because a rotary machine vibration diagnostic device that detects vibration abnormalities online can be provided at low cost.

Further, according to the next invention, a rotating machine vibration diagnostic apparatus is provided with a deterioration index storing means for storing a history value of a vibration deterioration index, and calculating a vibration deterioration index at an arbitrary time from the history value of the vibration deterioration index. If it is provided with the deterioration prediction means, it becomes possible to predict the vibration deterioration index at any time,
There is an effect that an inexpensive rotary machine vibration diagnosis device that can predict the time when the vibration level reaches the abnormal region and the dangerous region can be predicted.

Further, according to the next invention, the rotating machine vibration diagnostic apparatus is provided with a deterioration index storing means for storing a vibration deterioration index of a characteristic parameter for each vibration cause of the rotating machine, and the noise removing means outputs. If the tendency of the vibration deterioration index calculated from the frequency spectrum is compared with the vibration deterioration index of the characteristic parameter for each vibration cause, a rotating machine vibration diagnosis device capable of diagnosing the vibration cause can be provided at low cost. effective.

According to another aspect of the present invention, there is provided a rotating machine vibration diagnosing apparatus which includes data relating a specific noise frequency component value of a rotating machine and a vibration frequency component value corresponding to the frequency. The apparatus includes noise prediction means for predicting a noise frequency component value from the vibration frequency component value output from the removal means using related data, and abnormality diagnosis processing means for outputting a signal when the noise frequency component value reaches a predetermined value. In such a case, even at a site with high background noise, a noise level abnormality can be detected without installing a sound insulation device, and without using expensive equipment such as a sound insulation device, a rotation that can detect an abnormality that must be judged based on the noise level can be detected. There is an effect that the machine vibration diagnosis device can be provided at low cost.

According to another aspect of the present invention, there is provided a rotating machine vibration diagnostic apparatus comprising: a noise frequency component value storage unit for storing a history value of a predetermined noise frequency component value output by a noise prediction unit; If the apparatus is provided with deterioration prediction means for calculating a predetermined noise frequency component value at an arbitrary time from a history value of the value, the predetermined noise frequency component value at an arbitrary time is predicted. There is an effect that an inexpensive rotary machine vibration diagnostic device capable of predicting the time to reach the region can be provided at low cost.

Further, according to the next invention, the rotating machine vibration diagnostic apparatus is configured such that the noise removing means outputs a frequency spectrum obtained by averaging repeated frequency components in the frequency spectrum output by the sampling means. Since the random noise of the frequency spectrum is removed, there is an effect that a rotating machine vibration diagnostic apparatus capable of analyzing vibration with higher accuracy using a frequency spectrum with less noise is provided.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a rotating machine vibration diagnosis device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of noise prediction means of the rotating machine vibration diagnosis device according to one embodiment of the present invention.

FIG. 3 is a flowchart showing the operation of an abnormality diagnosis processing means of the rotating machine vibration diagnosis apparatus according to one embodiment of the present invention.

FIG. 4 is a flowchart showing the operation of a deterioration prediction means of the rotating machine vibration diagnosis apparatus according to one embodiment of the present invention.

FIG. 5 is an explanatory diagram showing a relationship between a temporal change of a vibration deterioration index used in a rotating machine vibration diagnosis apparatus according to one embodiment of the present invention and threshold values of an abnormal region and a dangerous region.

FIG. 6 is an explanatory diagram showing a relationship between a time-dependent change of a vibration deterioration index, a pass / fail judgment, and threshold values of an abnormal area and a dangerous area used in the rotating machine vibration diagnosis apparatus according to one embodiment of the present invention.

FIG. 7 is an explanatory diagram in a case where the change with time of the vibration deterioration index of FIG. 6 has an inflection point.

FIG. 8 is a configuration diagram of a conventional rotating machine vibration diagnosis device.

[Explanation of symbols]

 Reference Signs List 4 Vibrometer 5 Tachometer 8 Signal sampling means 14 Noise removal means 15 Noise prediction means 17 Degradation index processing means 18 Abnormality diagnosis processing means 19 Deterioration prediction means 20 Degradation index recording means 21 Degradation index data file

────────────────────────────────────────────────── ─── Continuation of the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) G01H 17/00 H02K 11/00 JICST file (JOIS)

Claims (6)

(57) [Claims] 1. A rotational speed detecting means for detecting a rotational speed of a rotating machine, a vibration detecting means for detecting vibration of the rotating machine, and a signal output by the vibration detecting means is sampled for a predetermined period to output a frequency spectrum. Signal removing means, a noise removing means for outputting a frequency spectrum obtained by removing noise from a frequency spectrum outputted from the signal sampling means, and the rotation number detecting means among the frequency spectrum outputted from the noise removing means. Deterioration index processing means for calculating a vibration deterioration index which is a ratio of a rotation synchronization frequency component value synchronized with the rotation speed and an integer multiple of this frequency component value, and abnormality diagnosis processing means for outputting a signal when the vibration deterioration index reaches a predetermined value A vibration diagnostic device for a rotating machine, comprising: 2. A deterioration index storage means for storing a history value of a vibration deterioration index, and deterioration prediction means for calculating a vibration deterioration index at an arbitrary time from the history value of the vibration deterioration index. The rotating machine vibration diagnostic device according to claim 1. 3. A deterioration index storing means for storing a vibration deterioration index of characteristic parameters for each vibration cause of the rotating machine, wherein the tendency of the vibration deterioration index calculated from the frequency spectrum outputted by the noise removing means and the vibration The rotating machine vibration diagnosis apparatus according to claim 1, wherein the vibration cause is diagnosed by comparing a characteristic deterioration parameter of each cause with a vibration deterioration index. 4. A rotating speed detecting means for detecting a rotating speed of a rotating machine, a vibration detecting means for detecting vibration of the rotating machine, and a signal output by the vibration detecting means is sampled for a predetermined period to output a frequency spectrum. Signal sampling means, a noise removal means for outputting a frequency spectrum obtained by removing noise from a frequency spectrum output by the signal sampling means, a specific noise frequency component value of the rotating machine, and a vibration frequency component corresponding to this frequency. Noise prediction means having data associated with the noise frequency component value and predicting a noise frequency component value from the vibration frequency component value output by the noise removal means using the relevant data; and A rotating machine vibration diagnosis device, comprising: abnormality diagnosis processing means for outputting a signal when the value becomes a value. 5. A noise frequency component value storage means for storing a history value of a predetermined noise frequency component value output from the noise prediction means, and the predetermined noise at an arbitrary time from the history value of the predetermined noise frequency component value. The rotating machine vibration diagnostic device according to claim 4, further comprising: deterioration prediction means for calculating a frequency component value. 6. The apparatus according to claim 1, wherein said noise removing means outputs a frequency spectrum obtained by averaging repeated frequency components in the frequency spectrum outputted by said sampling means. The rotating machine vibration diagnostic device according to any one of the above.