JP3145845B2 - Noise elimination method for bearing diagnosis device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically removing noise in a bearing diagnostic device for automatically detecting an abnormality in a bearing, and more particularly to a method for removing noise in a bearing diagnostic device suitable for low-speed bearings. About.

[0002]

2. Description of the Related Art It is well known that a conventional bearing diagnostic apparatus detects vibration acceleration of a bearing, and its detection signal shows an abnormal waveform, and when its level becomes high, it judges that the bearing is abnormal. I have.

Further, the present applicant has proposed a method of detecting an AE generated from a bearing in the event of an abnormality and processing the detection signal by an event counting method to diagnose a bearing abnormality, especially a bearing that rotates at a low speed. Japanese Patent Application No. 61-294215).

FIG. 4 shows a main circuit configuration of a bearing diagnostic apparatus using sound. In FIG. 4, an AE wave generated from a bearing abnormality is converted into an electric signal by an AE sensor 1, and an envelope component A (see FIG. 4) is extracted by an envelope detector 2. The time t exceeding the threshold value in the waveform represented by the envelope component is counted by the event counter 4 . The counted time t is compared with a threshold value in the normal / abnormality determination circuit 5, and if it exceeds the threshold value, it is determined that an abnormality has occurred. It is known that a sound is generated when a solid is deformed or broken, and a fine crack is formed in the solid, and in the process of growing, an elastic wave generated as the crack progresses ( Ultrasound) is called AE (acoustic emission).

[0005]

In such a bearing diagnostic apparatus, there still remains a problem to be solved in that the accuracy is influenced by the magnitude of the time constant of the envelope detector. This will be described more specifically. When the time constant of the envelope detector 2 is set to be large, the waveform of the output signal of the envelope detector 2 becomes a smooth signal waveform as shown by a symbol A-1 in FIG. On the other hand, if the time constant of the envelope detector 2 is set relatively small, the waveform of the output signal of the envelope detector 2 has peaks and valleys appearing on the envelope as shown by a symbol B-1 in FIG. .

[0006] Therefore, the envelope waveform is divided, not counting the correct number of events n ₁ to exceed the threshold value L ₂ many times. Therefore, the time constant of the envelope detector 2 is set to be large in order to increase the accuracy of detecting the abnormality of the bearing.

[0007] However, at the time of measurement on site, it comes from the outside due to an electrical whisker-like shape having a short duration exceeding the threshold as shown in FIG. Noise was sometimes an obstacle.
This noise has a high frequency component and overlaps with the frequency band used for bearing diagnosis, so that it cannot be simply removed by a filter.

For this reason, in the conventional method, a method is employed in which noise is removed with a length of time based on the envelope waveform.

However, if the time constant is set to a large value as described above, the waveform width of the noise signal becomes large, and it becomes difficult to discriminate the duration from the abnormal signal.

As a result, the data is counted by the event counter 4 and erroneous diagnosis may occur due to noise.

In view of the above, an object of the present invention is to provide a method for removing noise of a bearing diagnosis device suitable for a bearing diagnosis device which measures an AE and detects an abnormality of a bearing.

[0012]

SUMMARY OF THE INVENTION In order to achieve the above object, according to the present invention, a sound generated by a bearing is converted into an electric signal by a sensor, and an AE indicated by the converted electric signal is obtained.
In a noise elimination method of a bearing diagnosis device for detecting an abnormality of a bearing by detecting a wave, the electric signal is shaped into a pulse waveform by an envelope detector having a long time constant and a waveform shaping circuit having a short time constant. When the number of generated pulses is smaller than a predetermined number, the electrical signal at this time is determined to be noise, and the electrical signal is invalidated during a period determined as the noise signal.

[0013]

According to the present invention, an abnormal AE caused by a bearing abnormality
The signal has a sudden waveform as shown in FIG. 6, and the inventor has found that the noise signal due to electrical is in the form of a mustache (see FIG. 6). Using the characteristics, the measurement signal is pulsed, and the number of pulses within a certain time is counted to identify a noise signal and an abnormal signal having a sudden waveform.

[0014]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows a system configuration of a bearing diagnosis device to which the present invention is applied. In FIG. 1, the same parts as those in the conventional system in FIG.

In FIG. 1, a filter 10 removes low frequency components of an input signal. The output signal of the filter 10 is provided to the waveform shaping circuit 11 and the envelope detector 2. A comparator is used in the waveform shaping circuit 11 so that the level of the output signal of the filter 10 is higher than a predetermined first threshold value L1 (different from the threshold value described in the related art). A pulse signal COMP which is turned ON only is generated.

In the noise elimination method according to the present invention, a noise component is eliminated by matching the pulse signal COMP output from the waveform shaping circuit 11 with the output signal SIG of the envelope detector 2. The output signal of the noise elimination circuit 12 is supplied to the event counter 4, and thereafter, the same abnormality detection processing as that of the related art is performed.

FIG. 2 shows a circuit configuration of the noise removing circuit 12, and FIG. 3 shows a signal waveform in the circuit. The AND circuit 22 functions as a switching circuit, and only when the level of the output signal of the envelope detector 2 is equal to or higher than a predetermined level,
That is, the output pulse signal of the waveform shaping circuit 11 is supplied to the ring-down counter 23 only when an abnormal signal waveform that allows the mixing of the noise component occurs. The comparator 21 compares the level of the signal SIG with a second threshold L2 in order to detect the occurrence timing of the abnormal signal waveform.

The ring counter 23 counts down from the number of cancellations, and when the count value becomes 0, EVE
The NT signal is turned on . Here, the number of cancellations is a numerical value for identifying an abnormal signal generated due to a noise component and a bearing abnormality, and the number of pulses of the COMP signal generated within the continuation time when the envelope waveform exceeds the second threshold is If the number is smaller than the number of cancellations, the SIG signal during this period is determined to be a noise component.

The event counter 24 consists of the event number counter 4 and the normal / abnormal determination circuit 5 in FIG. 1
You . In the present embodiment, the presence or absence of a bearing abnormality is determined by counting the number of occurrences of the signal EVENT within a unit time.

The operation of the thus constructed bearing diagnosis device will now be described. When the measurement signal generated by the AE sensor 1 passes through the filter 10, the frequency component is removed by invalidation, and the AE raw waveform shown in FIG. 3 is obtained. The AE raw waveform shown in FIG. 3 is a waveform in which an abnormal signal and noise generated due to a breakage of a part of the bearing are mixed, and the waveform level is kept constant in a normal state. The AE raw waveform is shaped by the waveform shaping circuit 11 to obtain the COMP signal waveform in FIG.

On the other hand, as the output signal of the envelope detection circuit 2, the SIG signal waveform shown in FIG. 3 is obtained. The output signal from the comparator 21 is turned on for the durations t1 and t2 when the SIG signal waveform is equal to or higher than the predetermined level. For this reason,
During this time, the SIG signal is supplied to the ring-down counter 23. When the number of cancellations is set to, for example, 3, the down counter 23 subtracts (decrements) the numerical value 1 from the initial value 3 every time a pulse signal is input, and turns the EVENT signal ON when the count value becomes 0. AE in FIG.
If the raw waveform is an abnormal waveform that is not noise, the number of pulses of the COMP signal becomes 3 or more, and the EVENT signal is turned on.

On the other hand, in the COMP signal corresponding to the noise component in the AE raw waveform, the number of pulses generated within a predetermined time becomes smaller than the numerical value 3. Therefore, the count value of the ring-down counter 23 does not become 0, and the EVENT signal remains off (see FIG. 3 ). As a result, a noise signal component is removed from the EVENT signal. The event counter 4 counts the number of occurrences of the EVENT signal, and the normal / abnormality determination circuit 5 generates an abnormality occurrence signal when the count value reaches the number of abnormality determinations within the unit measurement time. This abnormality occurrence signal is output to the outside as an alarm (alarm) signal to notify occurrence of an abnormality in the bearing.

[0023]

As described above, according to the present invention,
Even when detecting an abnormal signal component of an impact waveform from the measurement signal, an abnormal signal component resulting from a bearing abnormality in the measurement signal,
Since the noise signal component can be discriminated, the noise signal can be removed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a system configuration according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a circuit configuration of a noise elimination circuit according to the embodiment of the present invention.

FIG. 3 is a waveform diagram showing a signal waveform according to the embodiment of the present invention.

FIG. 4 is a block diagram showing a system configuration of a conventional example.

FIG. 5 is a diagram showing a change in an envelope waveform depending on the magnitude of a time constant in a conventional noise elimination circuit.

FIG. 6 is a diagram showing an AE waveform of noise and an AE waveform generated at the time of abnormality.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 AE sensor 2 Envelope detector 3 Conventional noise elimination circuit 4 Event counter 5 Normal / abnormal judgment circuit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) G01H 17/00 F16C 19/52 G01M 13/04 G01N 29/14

Claims (1)

(57) [Claims] 1. A noise elimination method for a bearing diagnostic device, wherein a sound generated by a bearing is converted into an electric signal by a sensor, and an AE wave indicated by the converted electric signal is detected to detect an abnormality of the bearing. The electric signal has a long time constant envelope detector and a time constant.
When the counted number of generated pulses is smaller than a predetermined number, the electric signal at this time is determined to be noise, and the pulse is shaped during the period determined to be the noise signal. A noise removing method for a bearing diagnostic device, wherein the electrical signal is invalidated.