JP2001255241A - Bearing diagnostic device for rotating machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To heighten diagnostic accuracy of a bearing, and convert a detected vibration signal into a sound and output it, by detecting an actual operation frequency of a rotating machine from the vibration of the bearing, and by calculating a characteristic frequency based on the actual operation frequency. SOLUTION: This bearing diagnostic device 8 of the rotating machine is composed of a vibration detection means 17, a signal conversion recording means 20, a diagnostic means 21, a voice output means 23 and an information processing means 24. In the state where an induction motor 1 is driven, the vibration of a rolling bearing 2 is detected, and the detected vibration signal is converted into a wave type calibrated voice signal, and the actual operation frequency is detected from the calibrated voice signal, and the characteristic frequency is calculated from the actual operation frequency, and abnormality diagnosis of the rolling bearing 2 is executed by comparing the characteristic frequency with a frequency component of the calibrated voice signal. A vibration sound at vibration detection and a vibration sound obtained by cutting an optional frequency component from the vibration sound at vibration detection are outputted as a voice.

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 a bearing of a rotating machine which detects vibration of the bearing of the rotating machine and diagnoses an abnormality of the bearing.

[0002]

2. Description of the Related Art FIG. 16 shows a conventional structure of a motor bearing diagnosis device 102 for diagnosing an abnormality of a bearing 101 by detecting vibration of a bearing 101 of a rotating machine, for example, an induction motor (hereinafter simply referred to as a motor) 100. It shows. An acceleration sensor 103 is mounted on an upper portion of an outer peripheral surface of a housing (not shown) that accommodates a bearing 101 of the electric motor 100, and a signal line 104 derived from the acceleration sensor 103 is connected to a vibration signal input terminal of a vibration measuring device 105. It is connected to the. The vibration measuring device 105 is provided with a function of storing detected vibration data in an external storage medium 106 that can be attached to and detached from the vibration measuring device 105 according to a predetermined format.

In the vibration measuring device 105, the vibration of the bearing 101 is detected while the electric motor 100 is driven, and the presence / absence of an abnormality of the bearing 101 is simply diagnosed from the vibration signal (1).
Next diagnosis). If it is determined in the primary diagnosis that the bearing 101 is abnormal, the personal computer 107 in which the bearing diagnosis software is installed reads out the vibration signal recorded in the external storage medium 106, By performing a precise diagnosis (secondary diagnosis), an abnormal portion of the bearing 101 is specified.

[0004]

Conventionally, there has been no vibration measuring device dedicated to the motor bearing diagnosis device 102, and a general-purpose vibration measuring device 105 has been used. Was composed.
This general-purpose vibration measuring device 105 is generally provided with a large number of vibration signal input terminals, and is configured to be capable of simultaneously capturing a large number of vibration signals. The score is limited low (for example, 1024 data scores). Therefore, in the motor bearing diagnosis device 102 configured by the general-purpose vibration measurement device 105, the frequency resolution of the detected vibration signal is low, and the actual rotation speed of the motor 100 (hereinafter, this actual rotation speed is referred to as the actual operation frequency) ) Could not be detected accurately. Therefore, the characteristic frequency of the bearing used when diagnosing the bearing 101 (frequency detected only when there is an abnormality in the rolling bearing) is calculated based on the command frequency set in the driving device of the electric motor 100. I was

However, when the command frequency differs from the actual operation frequency due to the size of the load, as in an induction motor, an accurate bearing diagnosis is performed using the characteristic frequency calculated based on the command frequency. I couldn't do that. In addition, since the frequency resolution of the detected vibration signal is low, it is not possible to accurately detect a characteristic frequency existing in a low frequency range (for example, several tens [Hz] or less).

[0006] Conventionally, a motor bearing diagnostic device 1
It was not easy to convert the vibration data detected in step 02 into sound and listen to the sound (vibration sound) with a human ear, so that it was not easy to convert the vibration data into sound.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and therefore has as its object to detect an actual operating frequency of a rotating machine based on a bearing vibration signal detected while the rotating machine is driven. Based on the actual operating frequency, the characteristic frequency of the bearing is calculated, and by comparing the characteristic frequency with the frequency component of the detected vibration signal, the abnormality of the bearing is diagnosed and the detected vibration signal is converted into sound. An object of the present invention is to provide a bearing diagnosis device for a rotating machine that can convert and output the converted bearing.

[0008]

According to a first aspect of the present invention, there is provided a bearing diagnosis device for a rotating machine, wherein a vibration detecting means for detecting vibration, and a vibration signal detected by the vibration detecting means are converted into an audio signal and recorded. A signal conversion recording unit, and a frequency analysis process of the audio signal recorded by the signal conversion recording unit to determine an actual rotation speed of the rotating machine; a characteristic frequency of the bearing is determined based on the actual rotation speed; Diagnostic means for diagnosing the abnormality of the bearing by comparing the frequency component of the audio signal subjected to the frequency analysis processing, and audio output means for outputting audio based on the audio signal recorded by the signal conversion recording means. It is characterized by having.

According to such a configuration, the vibration of the bearing of the driven rotating machine is detected, and the voice signal converted from the detected vibration signal is subjected to frequency analysis processing.
The actual number of rotations of the rotating shaft of the rotating machine can be obtained. Moreover, since the characteristic frequency of the bearing can be calculated based on the actual rotational speed, the abnormality of the bearing can be accurately diagnosed by comparing the characteristic frequency with the frequency component of the audio signal subjected to the frequency analysis processing. can do. Also,
By converting the vibration signal into an audio signal in the signal conversion recording means, the sound signal can be easily output as sound, so that the human ear can always hear the vibration sound.

A bearing diagnostic device for a rotating machine according to claim 2 is
An information processing means for filtering an audio signal recorded by the signal conversion recording means is provided, and the audio signal filtered by the information processing means is output as audio from the audio output means. According to such a configuration, the sound signal recorded by the signal conversion recording means can be filtered and output as sound, so that a human ear can hear the vibration sound from which a specific frequency component has been removed. .

According to a third aspect of the present invention, there is provided a bearing diagnosis device for a rotating machine.
In the diagnostic means, the audio signal recorded by the signal conversion recording means is calibrated based on the pre-recorded calibration audio signal, and the peak of the waveform obtained by performing envelope processing on the calibrated audio signal. A simple diagnosis means for detecting the bearing value when the peak value reaches a predetermined acceleration, and a calibrated sound when the simple diagnosis means determines that the bearing is abnormal. The frequency of the signal is analyzed to determine the actual rotational speed of the rotating machine.
A precise diagnosis means for determining a bearing abnormal frequency by determining a characteristic frequency of the bearing based on the actual rotational speed and comparing the characteristic frequency with the frequency component of the audio signal subjected to the envelope processing. It is characterized by the following.

According to such a configuration, only when the simple diagnosis means detects an abnormality of the bearing, the precise diagnosis means identifies the abnormal part of the bearing, so that the diagnosis time of the bearing can be reduced. it can. Then, other factors considered to be abnormal can be removed.

According to a fourth aspect of the present invention, there is provided a bearing diagnostic device for a rotating machine.
A vibration signal recording means for recording and reproducing a vibration signal detected by the vibration detecting means is provided. According to such a configuration, the vibration signal of the bearing while the rotating machine is driven can be detected and recorded only by the vibration detection means and the vibration signal recording means. Accordingly, the vibration measurement of the bearings of a plurality of rotating machines can be performed simultaneously only by providing a plurality of vibration detecting means and vibration signal recording means without preparing a bearing diagnosis device for a plurality of rotating machines. Further, a large number of vibration signals and a large amount of vibration signals can be recorded. Thereby, even if the abnormal vibration generated in the bearing occurs intermittently or the reproducibility is low, the abnormal vibration sound can be reliably detected by continuously recording the vibration signal for a long time. .

According to a fifth aspect of the present invention, there is provided a bearing diagnosis device for a rotating machine.
The signal conversion recording means generates an audio signal by A / D converting the vibration signal at a sampling frequency based on a wave format, and records the audio signal in a wave format. According to such a configuration, a wave format A / A provided as standard equipment in a general-purpose Windows OS is used.
Since the audio signal can be generated using the D conversion processing function, there is no need to add new hardware or software, and the manufacturing cost of the bearing diagnosis device for the rotating machine can be reduced. Further, by converting the vibration signal into an audio signal, audio output processing can be easily performed in addition to the bearing diagnosis processing. In addition, since the file in the wave format can be recognized by a personal computer on which Windows OS is installed, it can be recognized by a general personal computer.
This audio signal can be easily heard as sound.

According to a sixth aspect of the present invention, there is provided a bearing diagnostic device for a rotating machine.
The signal conversion recording means captures two different vibration signals in a stereo manner, and performs A / D conversion of these two vibration signals at a sampling frequency based on the wave format to generate audio signals in parallel and simultaneously. It is characterized by performing. According to such a configuration,
The efficiency of the process of converting the vibration signal detected by the vibration detection means into an audio signal can be improved.

According to a seventh aspect of the present invention, there is provided a bearing diagnosis device for a rotating machine.
The diagnostic means determines that the number of data points of the fast Fourier transform performed at the time of the frequency analysis processing of the audio signal generated by A / D converting the vibration signal at the sampling frequency based on the wave form is set to 16384 points or more. Features. According to such a configuration, since the frequency resolution by the fast Fourier transform can be increased, the actual rotational speed of the rotating machine can be detected with high accuracy in the frequency analysis processing for diagnosing a bearing abnormality, and When comparing the characteristic frequency with the frequency component of the audio signal subjected to the frequency analysis processing, it is possible to accurately identify each characteristic frequency of the parts constituting the bearing, including the low frequency region.

A bearing diagnostic device for a rotating machine according to claim 8 is
The vibration detecting means comprises an acceleration sensor and an acceleration sensor amplifier. According to such a configuration, since the configuration is simple, portability is convenient.

According to a ninth aspect of the present invention, there is provided a bearing diagnosis device for a rotating machine,
The vibration signal recording means is constituted by a mini-disc recorder. According to such a configuration,
Easy to operate and convenient to carry.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [First Embodiment] A first embodiment in which the present invention is applied to an induction motor will be described below with reference to FIGS.

<Structure of Rolling Bearing and Diagnosis Method of Rolling Bearing Abnormality> First, the structure of the rolling bearing 2 mounted on an induction motor (hereinafter simply referred to as a motor) 1 (see FIG. 1) as a rotating machine is described. A method of diagnosing an abnormality such as damage or defect of the rolling bearing 2 by detecting vibration of the rolling bearing 2 while driving the electric motor 1 will be described.

As shown in FIG. 2, the rolling bearing 2 comprises an outer ring 3, an inner ring 4, a rolling element 5, a retainer 6, and small parts (not shown) such as an oil sheet. The symbols in the figure represent D: pitch circle diameter of the rolling bearing 2 d: diameter of the rolling element 5 r1: radius of the inner race 4 raceways r2: radius of the outer race 3 raceways α: contact angle.

When the vibration of the rolling bearing 2 is detected while the electric motor 1 is driven, when the rolling bearing 2 is normal, the vibration of the frequency component proportional to the rotation speed of the rotating shaft 7 is detected. . On the other hand, when the components of the rolling bearing 2 have abnormalities such as damage and defects, the vibration of the rolling bearing 2 is detected, and the vibration of the frequency component proportional to the rotation speed of the rotating shaft 7 (see FIG. 1) is obtained. Vibration of a frequency component different from the above is also detected. The frequency detected only when there is an abnormality in the rolling bearing 2 is called a characteristic frequency. This characteristic frequency indicates a different frequency depending on the abnormality of each component constituting the rolling bearing 2 and can be obtained by the following arithmetic expression. However, at this time, the rolling bearing 2 has (a) no sliding contact between the inner ring 4 and the outer ring 3 and the rolling element 5 (b) no deformation of each part when a radial or thrust load is applied. Assume that

(1) Characteristic frequency when the inner ring 4 has an abnormality

(Equation 1) (2) Characteristic frequency when the outer ring 3 has an abnormality

(Equation 2) (3) Characteristic frequency when the rolling element 5 has an abnormality

(Equation 3) (4) Characteristic frequency when the cage 6 has an abnormality

(Equation 4) Here, D: pitch circle diameter of rolling bearing 2 [mm] fr: rotation speed of rotating shaft 7 [rps] d: diameter of rolling element 5 [mm] Z: number of rolling elements 5 [integer] α: contact angle [ Radians].

Thus, while the electric motor 1 is driven, the frequency component of the vibration of the rolling bearing 2 is detected, and based on this frequency component, the actual rotational speed of the rotating shaft 7 (hereinafter, this actual rotational speed Frequency), calculate a characteristic frequency based on the actual operating frequency, and compare the calculated characteristic frequency with the frequency component of the detected vibration to determine whether the rolling bearing 2 has an abnormality. detection,
Then, the location of the abnormality is specified. For reference, FIG.
The relationship between various abnormalities of the rolling bearing 2 and the corresponding vibration waveforms will be described.

FIG. 3A shows a cycle of rotation of the rotating shaft 7. (B) shows a vibration waveform detected when the rolling bearing 2 is normal.
(C) shows a vibration waveform detected when poor lubrication of the oil sheet or uniform wear of the rolling bearing 2 occurs. (D) shows a vibration waveform detected when misalignment between the rotating shaft 7 and the partial bearing 2 occurs. (E), (f), (g), and (h) represent vibration waveforms detected when an abnormality occurs in the outer ring 3, the inner ring 4, the moving body 5, and the retainer 6, respectively.

<Structure of Motor Diagnosis Apparatus 8>
FIG. 1 shows a configuration of an electric motor bearing diagnosis device (hereinafter, simply referred to as a diagnosis device) 8. Cylindrical joining plates 11 and 12 are fixedly attached to the tip of the rotating shaft 7 of the electric motor 1 and the tip of the rotating shaft 10 of the load device 9, and the joining plates 11 and 12 are joined. The electric motor 1 and the load device 9 are joined by being fixed with bolts. As shown in FIG. 4, rolling bearings 2 for supporting a rotating shaft are mounted on both ends of the motor frame 13.
3 is housed in the housing 29.

An acceleration sensor 14 made of a piezoelectric element is attached to the upper portion of the outer peripheral surface of the housing 29 for accommodating these rolling bearings 2 by firmly bonding. When these acceleration sensors 14 are vibrated by the rotation of the rotating shaft 7 of the electric motor 1, the vibrations are converted into analog electric signals by the piezoelectric elements, and the signals derived from these acceleration sensors 14 are output as vibration signals. Line 15 is transmitted. These two signal lines 15 are connected to the L input terminal and the R input terminal of the acceleration sensor amplifier 16. The acceleration sensor amplifier 16 amplifies a vibration signal input from the L input terminal and the R input terminal to a predetermined level of amplitude and outputs the amplified signal from a stereo output terminal. The acceleration sensor 14 and the acceleration sensor amplifier 16 constitute a vibration detecting unit 17.

A stereo signal line 18 connected to an output terminal of the acceleration sensor amplifier 16 is connected to a microphone input terminal provided in a sound interface in a personal computer (hereinafter, simply referred to as a personal computer) 19. Although the personal computer 19 is shown in a block diagram for each configuration in FIG. 1, it is actually configured by software mainly including a CPU. This PC 1
Reference numeral 9 denotes a general-purpose device (not shown),
A ROM, a RAM, a hard disk, a floppy (registered trademark) disk drive, a keyboard, an input interface, an output interface, a sound interface, and the like are provided. As the OS, a Windows OS is installed. Then, the diagnostic program recorded in the ROM is written into the CPU, so that the CPU controls all operations of the personal computer 19.

The vibration signal input to the microphone input terminal is
It is input to the signal conversion recording means 20. In the signal conversion recording means 20, before the diagnosis of the rolling bearing 2 is performed, an audio signal for calibration (hereinafter, this audio signal is referred to as an audio signal for calibration) is previously stored in an external recording medium such as a hard disk or a floppy disk. (Hereinafter, the hard disk and the external recording medium are collectively referred to as a recording medium).
The calibration audio signal is a digital audio signal generated by subjecting an analog vibration signal to A / D conversion at a sampling frequency based on a wave format.

When the vibration detection of the rolling bearing 2 is started, the signal conversion and recording means 20 A / D converts the input analog vibration signal at a sampling frequency based on the wave format, and converts the analog vibration signal into a digital signal. An audio signal is generated, the audio signal is calibrated to a quantitative value representing the magnitude of the gravitational acceleration based on the calibration audio signal, and is recorded on a recording medium. Hereinafter, the calibrated audio signal is referred to as a calibrated audio signal.

When the diagnosing means 21 starts the diagnosing process, the calibrated audio signal is subjected to a frequency analysis process so that an abnormality of the rolling bearing 2 is diagnosed. Diagnostic means 2
Although not shown in FIG. 1, a simple diagnostic means and a precise diagnostic means are provided, and the simple diagnostic means performs a signal obtained by performing an envelope process on the calibrated audio signal (hereinafter, referred to as a "first").
This signal is referred to as an envelope-processed audio signal), and the peak value of the gravitational acceleration is detected. When the peak value reaches a predetermined gravitational acceleration, the rolling is performed such that the rolling bearing 2 is abnormal. It is determined whether or not the bearing 2 is abnormal. In the precise diagnostic means, when the simple diagnostic means determines that the rolling bearing 2 is abnormal, the actual operating frequency of the electric motor 1 is detected by performing frequency analysis processing on the calibrated audio signal, and based on the actual operating frequency. Then, the characteristic frequency of the rolling bearing 2 is calculated, and by comparing this characteristic frequency with the frequency component of the envelope-processed audio signal, an abnormal portion of the rolling bearing 2 is specified. The result diagnosed by the simple diagnostic means and the precise diagnostic means is output to the monitor 22.

The audio output means 23 outputs a calibrated audio signal,
In addition, an audio output of a signal obtained by subjecting the calibrated audio signal to the band cut filter processing (hereinafter, this signal is referred to as a band cut audio signal) is performed. When the calibrated audio signal is output as an audio signal, first, the calibrated audio signal recorded on the recording medium is written to the RAM. Then, the calibrated audio signal is output from the speaker as the vibration sound of the rolling bearing when the vibration is detected.

When the sound output of the band cutoff sound signal is performed, first, the band cutoff filter processing of the calibrated sound signal is performed in the information processing means 24 (described later). Then, in response to a band cutoff audio signal (described later) generated by the information processing means 24, the band cutoff audio signal is output from the speaker. This vibration sound is a vibration sound in which a vibration component in a certain frequency band is cut from the vibration sound of the rolling bearing 2 at the time of vibration detection.

In the information processing means 24, when the band cut filter processing of the calibrated audio signal is started, first, the calibrated audio signal recorded on the recording medium is written into the RAM. Then, based on the set frequency band, the band cut filter processing of the calibrated audio signal is performed, and the generated band cut audio signal is output to the audio output unit 23.

<Description of Operation of Diagnostic Apparatus 8> Next, the operation of the present embodiment will be described with reference to FIGS.
When the software for diagnosing the rolling bearing 2 is started on the personal computer 19, a main menu screen for diagnosing the rolling bearing 2 is displayed on the monitor 22, although not shown. On this main menu screen, the following processing items can be selected: • “Record calibration audio signal” • “Bearing diagnosis” • “Audio output of calibrated audio signal”

Each processing item on the main menu screen is executed by pressing a key on the keyboard corresponding to each processing item. When these keys are pressed, the CPU
Flags of execution processing items are set therein, and software processing as described below is executed according to these flags.

<Recording of calibration audio signal> First, in order to diagnose the rolling bearing 2, it is necessary to record the calibration audio signal on a recording medium. Therefore, the operation when the processing item of “recording of calibration audio signal” is selected from the main menu screen will be described. In the signal conversion recording means 20, a plurality of vibration signals of different gravitational accelerations are converted into calibration audio signals and recorded on a recording medium. When “Recording of calibration audio signal” is selected on the main menu screen, the monitor 22 displays
A menu screen for setting the gravitational acceleration of the vibration signal to be recorded as the calibration audio signal is displayed. In the present embodiment, the gravitational acceleration is changed from 1 [G] to 1 according to the menu screen.
Until 0 [G], the vibration signal at 1 [G] is set to be recorded as a calibration audio signal.

In the case of recording the calibration voice signal, although not shown, the vibration detecting means 17 as shown in FIG. 1 is connected to a vibrator or the like which can vibrate at an accurate and constant gravitational acceleration. . The acceleration sensor connected to the shaker is
One or two may be sufficient. Then, the vibrator is vibrated at a certain fixed gravitational acceleration. At this time, in the vibration detecting means 17, the vibration of the vibrator is converted into an analog electric signal, and the converted signal is recorded as a vibration signal.
Output to 0.

The signal converting and recording means 20 converts the input analog vibration signal into a maximum of four based on the wave form.
A / D conversion is performed at a sampling frequency of 4.1 [kHz], and a digital calibration audio signal is generated. And
This calibration audio signal is recorded on a recording medium as a calibration audio signal file in a wave format. In this embodiment, a calibration audio signal when the gravitational acceleration of the vibrator is vibrated every 1 [G] from 1 [G] to 10 [G] by the above method is recorded on the recording medium. . In FIG. 5, as an example, the gravitational acceleration of the shaker is set to 1 [G], 2 [G], 3 [G].
The waveform of the vibration signal detected by the vibration detecting means when the vibration is set at [G] is shown.

When the recording of the plurality of calibration voice signals is completed, the signal conversion recording means 20 detects the average value of the peaks of the calibration voice signals at the respective gravitational accelerations.
On the basis of the average value of these peaks, an approximate straight line that is linearly proportional to the gravitational acceleration and the average value of the peaks of the calibration audio signal (hereinafter, this approximate straight line is referred to as an approximate straight line for calibration) is calculated, It is recorded on a storage medium. FIG. 6 shows an example of calculating a calibration approximate straight line based on the average value of the peaks of a plurality of calibration audio signals.

<Bearing Diagnosis> When the processing item of “recording of calibration voice signal” is completed, “bearing diagnosis” can be executed. Therefore, the operation when the “Bearing diagnosis” processing item is selected from the main menu screen will be described with reference to FIGS.
This will be described with reference to the flowchart of FIG. The control of the transition between the steps in these flowcharts is performed by the CPU.

When "Bearing diagnosis" is selected on the main menu screen, as shown in FIG.
Then, the CPU selects whether to perform the diagnostic processing of the calibrated audio signal recorded on the recording medium or to perform the diagnostic processing by detecting the vibration of the rolling bearing 2 while the electric motor 1 is driven. Along with this, a menu screen for selecting and inputting these diagnostic processes is displayed on the monitor 22. Here, when a processing item for performing a diagnosis process by detecting vibration while the electric motor 1 is driven is selected, the process proceeds to step S2, where the vibration of the rolling bearing 2 is detected. Details of this vibration detection will be described with reference to the flowchart of FIG.

In step T1, a sampling frequency is set. Here, the sampling frequency based on the wave format is a maximum of 44.1 [kHz], the settable sampling frequency is 44.1 [kHz], and its dividing frequency (for example, 22.05 [kHz], 1
1.025 [kHz]). When the sampling frequency is input, the process proceeds to step T2. In the present embodiment, the sampling frequency is 44.1 [k
Hz] is set.

In step T2, the apparatus is in a state of waiting for input of a detection start key. At this time, the motor 1 is in a steady state based on the fact that the acceleration sensor 14 is fixedly mounted on the rolling bearing 2 of the motor 1 and the command frequency set in the driving device (not shown) of the motor 1. Check that it is driven by. If it is determined that the vibration detection of the rolling bearing 2 can be started, a detection start key is input, and the process proceeds to Step T3.

In step T3, the signal conversion recording means 20
In, based on the set sampling frequency,
A / D conversion of the vibration signal input from the vibration detection means 17 is performed, and an audio signal is generated. The vibration signals detected by the two acceleration sensors are input to the signal conversion and recording means 20 in a stereo system, and two different vibration signals are simultaneously processed in parallel to generate two sound signals. Is done. Then, control goes to a step T4. In step T4, data of the approximate straight line for calibration obtained in the processing item of “recording of audio signal for calibration” is read from the recording medium, and based on the approximate straight line for calibration, data of each sample of the audio signal is read out. The school moves to step T5.

In step T5, the calibrated audio signal is converted into wave format data, and recorded on a recording medium as a calibrated audio signal file. The two different vibration signals input in the stereo system are recorded on the recording medium as individual calibrated audio signal files. When the vibration detection is completed in this way, the process proceeds to step S4 in FIG. In step S1 of FIG.
If the diagnostic processing of the vibration signal recorded on the recording medium is selected, the process proceeds to step S3, where the calibrated audio signal file recorded on the recording medium is selected. Accordingly, a list of calibrated audio signal files recorded on the recording medium is displayed on the monitor. Then, when the calibrated audio signal file is selected, the process proceeds to step S4.

In step S4, the diagnostic means 21 reads out the data of the calibrated audio signal file for performing the diagnosis. If step S2 is executed in the previous step, the calibrated audio signal file recorded by the vibration detection is read out, and if step S3 is executed in the previous step, the selected audio signal file is read out. It is. Then, the read-out calibrated audio signal is written into the RAM, and the process proceeds to step S5.

In step S5, the diagnosing means 21 selects the rolling bearing 2. In the storage area provided in the diagnosis means 21, the dimensions of the components of the rolling bearing 2 necessary for calculating the characteristic frequency are arranged in advance according to the model number of the product and recorded as a rolling bearing database. Accordingly, a list of the rolling bearing database is displayed on the monitor 22, and the corresponding model number is selected. Here, if there is no corresponding model number in the rolling bearing database, the dimensions of the components of the rolling bearing 2 are individually input. The input dimensions of each part are additionally recorded in the rolling bearing database. When the input of the selection of the rolling bearing 2 is completed in this way, the process proceeds to step S6.

In step S6, numerical values required for diagnosis are input. In accordance with the menu screen on the monitor 22, as the numerical values necessary for the diagnosis, the command frequency set in the driving device of the electric motor 1 and the value of the gravitational acceleration for determining the abnormality of the rolling bearing 2 (hereinafter, referred to as (Referred to as “gravity acceleration for abnormality determination”) is input and recorded in a storage area provided in the diagnosis unit 21.
When the input of the numerical values necessary for the diagnosis is completed, the process proceeds to step S7.

In step S7, a simple bearing diagnosis for determining whether the rolling bearing 2 is abnormal is executed. The details of the bearing simple diagnosis will be described with reference to the flowchart of FIG. In step U1, the peak value of the gravitational acceleration is detected from the calibrated audio signal written in the RAM. The peak value is detected by comparing the values of all samples of the calibrated audio signal for each sample. Then, the detected peak value is stored in a storage area provided in the diagnosis unit 21. Then, control goes to a step U2.

In step U2, the peak value of the calibrated audio signal detected in step U1 is compared with step S6 in FIG.
Is compared with the value of the gravitational acceleration for abnormality determination set in the above.If the peak value of the calibrated audio signal is larger than the value of the gravitational acceleration for abnormality determination, it is determined that the rolling bearing has an abnormality. You. Then, control goes to a step S8 in FIG. In step S8, when it is determined that there is no abnormality in the rolling bearing 2 based on the result of the simple bearing diagnosis, the process proceeds to step S10. Rolling bearing 2
If it is determined that there is an abnormality, the process proceeds to step S9, and a bearing precision diagnosis is performed. The details of the bearing precision diagnosis will be described with reference to the flowchart of FIG.

In step V1, a fast Fourier transform of the calibrated audio signal written in the RAM is performed. Then, the process proceeds to step V2. In this fast Fourier transform processing, an operation is performed with a frequency resolution of about 2 [Hz] in order to detect the actual operating frequency of the electric motor 1 with high accuracy. In this embodiment, the number of data points for calculating the fast Fourier transform is 16384. At this time, the sampling frequency of the calibrated audio signal is 44.1.
Since it is A / D converted at [kHz], the frequency resolution by this fast Fourier transform processing is about 2.7.
[Hz].

FIG. 11 shows, by way of example, a frequency distribution waveform when a fast Fourier transform process is performed on a calibrated audio signal generated when the rolling bearing 2 having an abnormality in the outer ring 3 is diagnosed. As shown in FIG. 11, it can be seen that the frequency component of the calibrated audio signal has a plurality of portions where the amplitude of the gravitational acceleration is large in a specific frequency region. Hereinafter, these frequency regions are referred to as local maximum regions. These local maxima are the actual operating frequency of the electric motor 1, the characteristic frequencies of the components for calibrating the rolling bearing 2, and harmonic components of these frequencies.

In step V2, the actual operating frequency of the motor 1 is detected from the frequency components of the calibrated audio signal subjected to the fast Fourier transform. First, step S in FIG.
The command frequency value input in 6 is read from the storage area. Then, a maximum region closest to the value of the command frequency is detected from a plurality of maximum regions in the frequency component of the calibrated audio signal, and the frequency indicating the maximum value of the maximum region is the actual operating frequency of the electric motor 1. Is detected as
Thus, the detection of the actual operating frequency is performed, and the process proceeds to step V3.

In step V3, the characteristic frequency of the rolling bearing 2 is calculated based on the detected actual operating frequency and the dimensions of the component for calibrating the rolling bearing 2 selected in step S5 in FIG. Then, control goes to a step V4. In step V4, envelope processing of the calibrated audio signal is performed to remove harmonic noise components of the calibrated audio signal. Then, control goes to a step V5.

In step V5, a fast Fourier transform of the envelope-processed audio signal is performed. At this time, as in step U1 in FIG. 9, the number of data points of the envelope audio signal is set to 16384, and the calculation of the fast Fourier transform is performed. The reason is that, among the rolling bearings 2, the characteristic frequencies of the rolling elements 5 and the cage 6 are both in a low frequency range (for example, a frequency range of 10 [Hz] to 100 [Hz]).
(For example, at intervals of 10 [Hz]),
This is because the frequency resolution of the fast Fourier transform needs to be about 2 [Hz] in order to clearly identify these characteristic frequencies. In this embodiment, the envelope processed audio signal is A / D at a sampling frequency of 44.1 [kHz].
Since the converted calibrated audio signal has been subjected to envelope processing, the number of data points of the envelope-processed audio signal is 16
By performing fast Fourier transform on 384 points,
The frequency resolution is about 2.7 [Hz].

FIG. 12 shows, as an example, a frequency distribution waveform when the envelope-processed audio signal obtained when the rolling bearing 2 having an abnormality in the outer ring 3 is diagnosed is subjected to Fourier transform processing. As shown in FIG. 12, it can be seen that the actual operating frequency of the electric motor 1, the characteristic frequency of the outer ring 3, and the harmonic components of these frequencies are detected. Then, control goes to a step V6.

At step V6, the characteristic frequency of the rolling bearing 2 is compared with the frequency component of the envelope-processed audio signal. First, as shown in FIG. 12, in the frequency components of the envelope-processed audio signal, the value of the gravitational acceleration is detected at a frequency that matches each characteristic frequency. Next, a characteristic frequency having the largest value among the detected values of the gravitational acceleration is detected. Then, the components of the rolling bearing 2 corresponding to the detected characteristic frequency are:
It is specified as a main factor of the abnormality of the rolling bearing 2. FIG.
In the example shown in FIG. 2, since the value of the gravitational acceleration at the characteristic frequency when the outer ring 3 is damaged is the largest among the values of the gravitational acceleration at the respective characteristic frequencies, it is specified that the outer ring 3 is abnormal. Then, control goes to a step S10 in FIG.
In step S10, upon receiving the diagnosis results of the simple bearing diagnosis and the precision bearing diagnosis, the diagnosis results are displayed on the monitor 22 and recorded on a recording medium.

<Speech output of calibrated speech signal> Diagnosis device 8
Is configured to be able to output a vibration sound at the time of vibration detection as a sound based on data of a calibrated audio signal file recorded on a recording medium. Also,
A signal (band cut audio signal) generated by performing band cut filter processing for cutting a signal component in an arbitrary frequency region of the calibrated audio signal is configured to be output as audio. . Therefore, the operation when the processing item of “vibration signal sound output” is selected from the main menu screen will be described with reference to the flowchart of FIG.

First, at step W1, a calibrated audio signal file name for outputting audio is input. Along with this, a menu screen for inputting the file name is displayed on the monitor 22. When the calibrated audio signal file name is input, the data of the calibrated audio signal file corresponding to this file name is read from the recording medium and written to the RAM. Then, control goes to a step W2.

In step W2, a selection is made as to whether to output the vibration sound at the time of detecting the vibration or to output the vibration sound obtained by subjecting the calibrated audio signal to the band cut filter processing. Along with this, a menu screen for selecting these is displayed on the monitor 22. If the output of the vibration sound at the time of detecting the vibration is selected, the process proceeds to step W3, and if the output of the vibration sound obtained by performing the band cut filter processing on the calibrated audio signal is selected, the process proceeds to step W4.

In step W3, the sound reproduction processing of the calibrated sound signal is performed, so that the vibration sound at the time of detecting the vibration is output from the speaker. In step W4, a fast Fourier transform of the calibrated audio signal is performed. The fast Fourier transform here is also performed in step V1 or step V5 in FIG.
As in the fast Fourier transform process performed in the above, the calculation is performed with the number of data points being 16384 points. Then, control goes to a step W5.

At step W5, a frequency band for performing band cut filter processing of the calibrated audio signal is set. Along with this, a menu screen for inputting the value of the frequency band is displayed on the monitor 22.
When the setting of the frequency band is performed, step W6
Move to In step W6, based on the set frequency band, a band rejection filter process is performed on the calibrated audio signal that has been subjected to the fast Fourier transform process. Then, control goes to a step W7.

In step W7, an inverse fast Fourier transform process is performed on the calibrated audio signal that has been subjected to the band rejection filter processing, and is again converted into a time component calibrated audio signal.
Then, control goes to a step W8. In step W8,
In order to ensure the continuity of the sound at the time of the sound reproduction, a window function process is performed on the calibrated sound signal that has been subjected to the inverse fast Fourier transform processing, and a band cutoff sound signal is generated. Then, control goes to a step W9. In step W9, by performing the sound reproduction processing of the band cutoff sound signal subjected to the window function processing in step W8, the vibration sound in which the signal component in an arbitrary frequency region is cut off from the speaker 25 in the vibration sound at the time of detecting the vibration. Is output.

As described above, according to the present embodiment, while the electric motor 1 is driven, the vibration detecting means 17 detects the vibration of the rolling bearing 2, and the signal conversion recording means 20 outputs the detected vibration signal. Is converted into a calibrated audio signal in a wave format and recorded on a recording medium. Then, based on the recorded calibrated audio signal, the simple diagnostic means detects the presence or absence of an abnormality in the rolling bearing 2. If an abnormality is detected, the precision diagnostic means converts the calibrated audio signal into a precise diagnostic means. The frequency analysis processing detects the actual operating frequency of the electric motor 1, calculates the characteristic frequency of the rolling bearing 2 based on the actual operating frequency, and compares the characteristic frequency with the frequency component of the calibrated audio signal. Thus, an abnormal portion of the rolling bearing 2 is specified. Further, the sound output unit 23 can output a vibration sound at the time of vibration detection and a vibration sound obtained by cutting a signal component of an arbitrary frequency region in the vibration sound at the time of vibration detection.

According to such a configuration, since the actual operating frequency of the electric motor 1 is detected and the characteristic frequency of the rolling bearing 2 is calculated, an abnormal portion of the rolling bearing 2 can be accurately specified. it can. Therefore, even in the case of an induction motor or the like in which the command frequency and the actual operation frequency are different from each other, such as the electric motor 1 used in the present embodiment, it is possible to accurately specify an abnormal portion of the rolling bearing 2.

The diagnosis of the rolling bearing 2 is divided into a simple diagnosis and a precise diagnosis.
The presence or absence of the abnormality is detected, and the precise diagnosis is not performed for the one without abnormality, so that the diagnosis time can be reduced. In addition, the number of data points of the fast Fourier transform processing performed when the frequency analysis processing is performed on the calibrated audio signal A / D-converted based on the wave format having the maximum sampling frequency of 44.1 [kHz] is set to 16384 points or more. Since the frequency resolution can be made smaller than about 2 [Hz], the actual operation frequency can be detected with high accuracy.
The characteristic frequency of each component of the bearing, including the low frequency region, can be accurately identified.

Further, since the diagnostic device 8 can be constituted only by the vibration detecting means 17 and the personal computer 19, the manufacturing cost can be reduced. Moreover, by using a personal computer on which a general-purpose Windows OS is installed, an A / D conversion function at a sampling frequency based on a wave format provided as a standard feature in the Windows OS can be used. It is not necessary to newly add it, and the manufacturing cost can be further reduced. Further, since the bearing is easy to carry, the diagnosis of the rolling bearing 2 can be performed efficiently.

Since the signal conversion and recording means 20 captures two different vibration signals in a stereo system, the two vibration signals detected by the vibration detection means 17 are converted at a sampling frequency based on the wave format. A /
The processes for D-conversion can be performed simultaneously in parallel.
Further, the signal conversion recording means 20 converts the vibration signal detected by the vibration detection means 17 into an audio signal based on the wave format, so that the vibration sound at the time of the vibration detection can be easily output as sound. it can. In addition, since the file in the wave format can be confirmed on a personal computer on which the Windows OS is installed, the sound signal can be heard and heard on a general personal computer.

Further, by subjecting the calibrated audio signal to band cut filter processing, it is possible to reproduce a vibration sound obtained by cutting an arbitrary frequency component from the vibration sound at the time of vibration detection. Thereby, for example, the rolling bearing 2 of the electric motor 1 that generates abnormal vibration noise is diagnosed, and when an abnormal location of the rolling bearing 2 can be identified, the characteristic frequency component of the calibrated audio signal corresponding to the abnormal location is determined. By converting the vibration sound cut by the band rejection filter processing into a voice, it is possible to confirm the cause of the abnormality of the vibration sound with a human ear.

[Second Embodiment] Next, a second embodiment of the present invention will be described with reference to FIG. The same parts as those in FIG. 1 are denoted by the same reference numerals, and description thereof will be omitted. Only different parts will be described below. The stereo signal line 18 connected to the output terminal of the acceleration sensor amplifier 16 is connected to a microphone input terminal of a mini-disc recorder (hereinafter, simply referred to as MD recorder) 26 as a vibration signal recording unit. The stereo signal line 27 connected to the line output terminal of the MD recorder is connected to a personal computer 19.
It is connected to the microphone input terminal provided in the sound interface inside. Thus, the diagnostic device 8 is configured.

In the diagnostic device 8 of this embodiment, the vibration signal output from the vibration detecting means 17 is once recorded by the MD recorder 26. Then, by reproducing the vibration signal recorded on the MD recorder 26, the vibration signal output from the MD recorder 26 is input to the signal conversion recording means 20 of the personal computer 19, and the rolling bearing 2 is diagnosed. .

In this embodiment, “recording of calibration audio signal”
Is performed, the audio signal for calibration output from the vibration detecting means 17 is once recorded by the MD recorder 26. Then, the vibration signal for calibration recorded in the MD recorder 26 is reproduced and input to the signal conversion recording means 20 to be converted into a calibration audio signal, and the calibration audio signal is recorded on a recording medium. To

According to such a configuration, the vibration detecting means 1
7 and the MD recorder 26 alone can record the vibration signal of the rolling bearing 2 while the electric motor 1 is driven. Thus, when the vibration measurement of the rolling bearings 2 of the plurality of electric motors 2 is performed at the same time, a plurality of the vibration detection means 17 and the MD recorder 26 need only be prepared without preparing the plurality of diagnostic devices 8. Also, by using the MD recorder 26, a large number of vibration signals and a large-capacity vibration signal can be recorded, so that the abnormal vibration generated in the rolling bearing 2 occurs intermittently or has poor reproducibility. Even if it is low, abnormal vibration noise can be reliably detected by continuously recording the vibration signal for a long time.

Further, the MD recorder 26 is connected to the personal computer 19.
And by reproducing the vibration signal recorded on the MD recorder 26, the diagnosis of the rolling bearing 2 can be performed. Therefore, the vibration detection of the rolling bearing 2 and the diagnosis of the rolling bearing 2 can be performed in different places. The operation can be performed in a convenient time zone, and the working efficiency can be improved.

[Third Embodiment] Next, a third embodiment of the present invention will be described with reference to FIG. FIG.
The same parts as those described above are denoted by the same reference numerals and description thereof will be omitted, and only different parts will be described below. One acceleration sensor 14 is firmly bonded and attached to an upper portion of an outer peripheral surface of a housing 29 that accommodates the rolling bearing 2 on the side where the electric motor 1 and the additional device 9 are joined. A signal line 15 derived from the acceleration sensor 14 is connected to an input terminal on one side of an amplifier 16 for an acceleration sensor. A microphone 28 capable of detecting a human voice or the like is connected to the other input terminal of the acceleration sensor amplifier 16. Thus, the diagnostic device 8 is configured.

In the diagnostic device 8 of this embodiment, the vibration signal output from the vibration detecting means 17 is once recorded by the MD recorder 26. Then, by reproducing the vibration signal recorded on the MD recorder 26, the vibration signal output from the MD recorder 26 is input to the signal conversion recording means 20 of the personal computer 19, and is converted into a calibrated audio signal. It is recorded on a recording medium. At this time, microphone 2
The human voice detected by step 8 is output as voice by selecting "voice output of calibrated voice signal" and selecting a calibrated voice signal file in which this human voice is recorded.

In this embodiment, “recording of calibration audio signal”
Is performed, the audio signal for calibration output from the acceleration sensor 14 is once recorded by the MD recorder 26. Then, the vibration signal for calibration recorded in the MD recorder 26 is reproduced and input to the signal conversion recording means 20 to be converted into a calibration audio signal, and the calibration audio signal is recorded on a recording medium. To According to such a configuration, the test information such as the test conditions of the diagnosis, the test date, and the test place are recorded in the MD recorder 26 by the human voice using the microphone 28 and the personal computer 19.
Since it can be recorded as an audio signal, the form of recording paper and writing utensils becomes unnecessary. Also, since the test information and the vibration signal detected by the diagnosis are always recorded on the MD recorder as a pair of data, the recording paper on which the test information is written may be lost as in the case of recording on paper, or the test may be lost. It is possible to prevent the correspondence between the information and the vibration signal detected by the diagnosis from being mistaken.

The present invention is not limited to the embodiment described above and shown in the drawings.
Extension is possible. In the first embodiment of the present invention, two acceleration sensors 14 are connected to the acceleration sensor amplifier 16. However, as in the third embodiment, one acceleration sensor 14 is connected.
And the microphone 28 may be connected to the other side.

In the first and second embodiments of the present invention, the two acceleration sensors 14 are connected to the acceleration sensor amplifier 16, but may be one. In the embodiment of the present invention, the acceleration sensor 14 made of a piezoelectric element is used, but any acceleration sensor that converts vibration into an electric signal may be used.

In the embodiment of the present invention, the band rejection filter processing is performed in the information processing means 24.
You may make it perform a low-pass filter process, a high-pass filter process, a band-pass filter process, etc. In the embodiment of the present invention, the stereo signal line 18 or 27 is connected to the microphone input terminal of the personal computer 19 and the MD recorder 26, but may be connected to the line input terminal.

In the embodiment of the present invention, the diagnosis of the rolling bearing 2 is performed. However, the diagnosis of the sliding bearing can be performed by incorporating the arithmetic expression of the characteristic frequency of the sliding bearing into the software of the personal computer 19. In the embodiment of the present invention, the diagnosis of the bearing of the induction motor is performed. However, the diagnosis of the bearing of the rotating machine in general, such as an AC motor or a DC motor, may be performed. In the embodiment of the present invention, the audio signal is generated based on the wave format provided as standard equipment in the Windows OS.
It may be applied to other sound file formats such as the AIFF format.

In the embodiment of the present invention, the information processing means 24 is provided in the diagnostic apparatus 8 so that the band cut filter processing of the calibrated audio signal is performed. Excluding this, the diagnostic measure 8 may be configured. The audio output means is a Windows OS
For example, software for sound reproduction which is provided as a standard feature in, for example, may be used.

In the embodiment of the present invention, the diagnosis means 21 includes:
The simple diagnosis means and the precision diagnosis means are provided, and when the simple diagnosis means determines that there is an abnormality in the bearing, the precision diagnosis means performs a diagnosis for specifying an abnormal portion of the bearing. However, the present invention is not limited to this. For example, after determining whether or not there is an abnormality in the bearing by the simple diagnosis means, it may be possible to select whether or not to perform the diagnosis by the precision diagnosis means.

[0085]

As is apparent from the above description, the bearing diagnosis apparatus for a rotating machine according to the present invention detects the actual operating frequency of the rotating machine by performing frequency analysis processing on the vibration signal of the bearing. The characteristic frequency of the bearing is calculated on the basis of the above, so even if the command frequency and the actual operating frequency are different from each other as in the case of an induction motor, accurate bearing diagnosis can be performed. This provides an excellent effect that a vibration sound and a vibration sound obtained by cutting an arbitrary frequency component in the vibration sound at the time of vibration detection can be easily output as a sound.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a bearing diagnosis device for a rotating machine according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a rolling bearing.

FIG. 3 is a diagram showing a relationship between various abnormalities of a rolling bearing and a vibration waveform.

FIG. 4 is a cross-sectional view of a rolling bearing of an electric motor.

FIG. 5 is a waveform diagram of a vibration signal detected by vibration for calibration.

FIG. 6 is a diagram showing a calculation example of an approximate straight line for calibration.

FIG. 7 is a flowchart showing an operation of diagnosing a rolling bearing.

FIG. 8 is a flowchart showing the operation of vibration detection.

FIG. 9 is a flowchart showing the operation of a simple diagnosis of a rolling bearing.

FIG. 10 is a flowchart showing the operation of precision diagnosis of a rolling bearing.

FIG. 11 is a waveform diagram showing an example of fast Fourier transform processing of a calibrated audio signal.

FIG. 12 is a waveform diagram showing an example of fast Fourier transform processing of an envelope-processed audio signal.

FIG. 13 is a flowchart showing the operation of audio output.

FIG. 14 is a diagram corresponding to FIG. 1, showing a second embodiment of the present embodiment;

FIG. 15 is a view corresponding to FIG. 1, showing a third embodiment of the present embodiment.

FIG. 16 is a diagram corresponding to FIG. 1 showing a conventional example.

[Explanation of symbols]

In the drawings, 1 is an induction motor (rotating machine), 2 is a rolling bearing, 3 is an outer ring, 4 is an inner ring, 5 is a rolling element, 6 is a cage, 7
Is a rotating shaft, 8 is a diagnostic device (bearing diagnostic device for a rotating machine), 1
4 is an acceleration sensor (vibration detecting means), 16 is an acceleration sensor amplifier (vibration detecting means), 17 is vibration detecting means, 1
9 is a personal computer, 20 is a signal converter recording means, 21 is a diagnostic means, 22 is a monitor, 23 is an audio output means, 24 is an information processing means, 25 is a speaker, 26 is a mini disk recorder, and 28 is a microphone.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Keiichi Abe 1-1-1, Shibaura, Minato-ku, Tokyo Toshiba Corporation Headquarters (72) Inventor Hisanori Shibata 2121 No. 2 Nao, Asahimachi, Mie-gun, Mie Prefecture Inside the Toshiba Mie Plant (72) Inventor Fumiaki Takeuchi 2121 Nagoya, Asahi-machi, Mie-gun, Mie Prefecture F-term in the Toshiba Mie Plant (reference) BA02 CC43 DD05 5H611 AA01 BB01 PP03 QQ09 UA04

Claims (9)

[Claims] 1. A vibration detecting means for detecting vibration of a bearing of a rotating machine, a signal converting and recording means for converting a vibration signal detected by the vibration detecting means into an audio signal and recording the signal, and a signal converting and recording means The actual speed of the rotating machine is obtained by performing frequency analysis processing on the recorded audio signal, and the characteristic frequency of the bearing is obtained based on the actual rotation number. The characteristic frequency and the frequency component of the audio signal subjected to the frequency analysis processing are obtained. Diagnosing means for diagnosing an abnormality of the bearing by comparing with the above, and audio output means for outputting an audio based on the audio signal recorded by the signal conversion recording means, apparatus. 2. An information processing means for filtering an audio signal recorded by a signal conversion recording means, wherein the audio signal filtered by the information processing means is output as audio from an audio output means. The bearing diagnosis device for a rotating machine according to claim 1. 3. The diagnostic means is obtained by calibrating an audio signal recorded by a signal conversion recording means based on a previously recorded calibration audio signal, and subjecting the calibrated audio signal to envelope processing. A simple diagnostic means for detecting a peak value of the detected waveform and determining that there is a bearing abnormality when the peak value reaches a predetermined acceleration; and The frequency of the calibrated audio signal is subjected to frequency analysis processing to determine the actual rotational speed of the rotating machine, the characteristic frequency of the bearing is determined based on the actual rotational speed, and the characteristic frequency and the frequency component of the audio signal subjected to the envelope processing are obtained. 2. The bearing diagnosis device for a rotating machine according to claim 1, further comprising: a precision diagnosis unit configured to identify an abnormal portion of the bearing by comparing with the above. 4. The bearing diagnostic device for a rotating machine according to claim 1, further comprising vibration signal recording means for recording and reproducing a vibration signal detected by the vibration detection means. 5. The signal conversion and recording means generates an audio signal by A / D converting a vibration signal at a sampling frequency based on a wave format, and records the audio signal in a wave format. Item 3. A bearing diagnosis device for a rotating machine according to item 1. 6. The signal conversion and recording means captures two different vibration signals in a stereo manner, thereby obtaining these two vibration signals.
2. The bearing diagnosis apparatus for a rotating machine according to claim 1, wherein the processing of generating an audio signal by A / D converting two vibration signals at a sampling frequency based on a wave format is performed simultaneously in parallel. 7. The diagnostic means reduces the number of data points of the fast Fourier transform performed in the frequency analysis processing of the audio signal generated by performing the A / D conversion on the vibration signal at the sampling frequency based on the wave form to 16384 points. The bearing diagnosis device for a rotating machine according to claim 5, wherein: 8. The bearing diagnosis device for a rotating machine according to claim 1, wherein said vibration detecting means comprises an acceleration sensor and an amplifier for said acceleration sensor. 9. The bearing diagnostic device for a rotating machine according to claim 4, wherein said vibration signal recording means is constituted by a mini-disc recorder.