KR101482509B1 - Diagnosis System and Method of Bearing Defect - Google Patents

Abstract

The present invention relates to a system and a method for diagnosing a bearing defect by analyzing complexity and energy concentration of signals based on vibration signals generated from a bearing in rotary equipment. The system for diagnosing a defect in a bearing of rotary equipment comprises: a sensor unit which measures a vibration signal of the bearing; a signal processing unit which includes a filter which extracts a signal of natural frequency band from the vibration signal, and a signal processing device which calculates a complexity degree and an energy concentration ratio from the vibration signal after filtering and the vibration signal before filtering; and display unit which includes a display device displaying whether the bearing has the defect and an alarm device generating an alarm if the bearing has the defect.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a bearing defect diagnosis system,

The present invention relates to a method of diagnosing a bearing defect, and more particularly, to a system and method for diagnosing a bearing defect by analyzing signal complexity and energy concentration based on a vibration signal generated in a bearing of a rotating facility.

In general, Rolling Element Bearing (Rolling Element Bearing) is a basic component required for the operation of various machinery. It can be subjected to high pressure during operation and cause mechanical failure even with small surface defects. These bearings are used as an essential element in various industrial machinery, and the first anomaly occurs when the machinery is operated under abnormal conditions. Therefore, continuous monitoring of the condition of the bearings is required for stable operation of the mechanical device, and a recent Condition Monitoring System essentially includes a function for monitoring and diagnosing the defects of the bearings.

The most common and effective way to diagnose bearings defects is to use the envelope spectrum method to obtain the envelope from the vibration signal of the bearing and analyze the frequency through FFT (Fast Fourier Transform). This method can be compared with the defect frequency determined according to the specification of the bearing, which is a highly reliable method, and a new method may be derived by combining with other signal processing methods (see Prior Art 1).

However, in the above method, the frequency spectrum obtained from the actually measured vibration signal is influenced by factors such as (i) resolution used in the frequency analysis process, (ii) calculation error occurring in the frequency analysis process, and (iii) nonlinearity included in the vibration signal The failure frequency calculated theoretically from the specification of the bearing does not coincide with the defect frequency. If these factors are taken into consideration, reliability of bearing diagnosis by frequency spectrum can be secured. For example, using high-specification data analysis hardware that can handle large amounts of data to increase frequency resolution can improve accuracy when compared to the theoretical calculated fault frequencies.

On the other hand, general bay fault monitoring is performed by trend monitoring based on accumulated vibration data through continuous monitoring. That is, the magnitude of the vibration data at the steady state with respect to the bearing and the magnitude of the vibration data to be newly measured are compared, and when the difference is greater than or equal to a predetermined magnitude, it is determined that a defect exists in the bearing. This method can be applied only if there is vibration data accumulated for a certain period of time. If the same facility is installed at another position, vibration characteristics are different. Therefore, . As a result, the above-described method needs to individually manage the vibration monitoring standard customized for each equipment, which causes a problem in that all costs associated with the maintenance and repair of the equipment are increased.

Therefore, in order to monitor the defects of the bearings more efficiently, it is necessary to (1) require no comparison with the defect frequency, (2) do not require high-specification hardware for the monitoring of bearing defects, (3) ④ It is required to build a system with low cost through this.

[Prior Art 1] United States Patent No. 7,602,985 B2

SUMMARY OF THE INVENTION The present invention has been made in order to solve the above-mentioned problems, and it is an object of the present invention to provide a method and apparatus for detecting defects of a bearing from a vibration signal measured by a sensor, And a system for diagnosing bearing defects that can be performed by a system.

It is another object of the present invention to provide a system and method for diagnosing bearing defects which can improve the efficiency of maintenance of equipments without requiring steady state vibration data in the process of determining the presence of defects of the bearings .

According to an aspect of the present invention, there is provided a system for diagnosing a defect occurring in a bearing of a rotating facility, the system comprising: a sensor unit for measuring a vibration signal of a bearing; A filter for extracting a natural frequency band signal with respect to the measured vibration signal and a complexity degree of energy concentration and a signal from a vibration signal subjected to the filtering process and a vibration signal before the filtering process, A signal processing unit for extracting a monitoring factor value and comparing the extracted monitoring factor value with a preset reference value to determine the presence or absence of a defect in the bearing; And a display unit including a display device for indicating the presence or absence of a defect in the bearing and an alarm device for generating an alarm when a defect is generated in the bearing.

According to another aspect of the present invention, there is provided a method of diagnosing a defect occurring in a bearing of a rotating facility, the method comprising the steps of: (a) measuring a vibration signal of a bearing in a sensor unit installed in a housing of a bearing; (b) a signal processing step of acquiring and filtering the vibration signal measured in the step (a) to obtain the energy concentration Er and the complexity C of the signal and calculating a monitoring factor Sb therefrom; And (c) determining a state of the bearing by comparing the monitoring factor Sb of the step (b) with a reference value Ref.

Here, the energy concentration Er is defined as a ratio of the frequency spectrum energy sum of the interest band and the entire frequency band, the complexity C of the signal is defined as the complexity obtained from the interest band vibration signal, (Sb) is a parameter defined as a dimensionless value between 0 and 1 from the energy concentration (Er) and the complexity (C) of the signal.

In the step (c), the reference value Ref is set to a value between 0.1 and 0.2. When the monitoring factor Sb exceeds the reference value Ref, it is determined that there is a defect in the bearing .

Further, the method for diagnosing bearing defects may further include: (d) generating an alarm when a failure is detected in the bearing in step (c).

The present invention having such a configuration can be applied easily and conveniently to any facility since it uses only the vibration signal measured in real time in the present time as judgment data on the presence or absence of bearing defects and does not need the accumulated data required in general tendency monitoring There is an advantage.

In addition, since the present invention defines a monitoring factor indicating a defect state to determine whether there is a bearing defect, it is not necessary to compare with a theoretical calculated defect frequency, and the system can be configured with simple hardware. That is, it does not require a high-end frequency analyzer for obtaining a high frequency resolution, and thus is cost-effective.

1 is a block diagram showing a configuration of a bearing defect diagnosis system according to an embodiment of the present invention;
2 is a flowchart illustrating a process for diagnosing bearing defects according to an embodiment of the present invention.
3 is a graph showing the frequency spectrum of the steady state in the bearing defect diagnosis system of the present invention,
4 is a graph showing the frequency spectrum of the inner ring defect state in the bearing defect diagnosis system of the present invention, and Fig.
5 is a graph showing a frequency spectrum of an outer ring defect state in the bearing defect diagnosis system of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features and advantages of the present invention will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram illustrating a configuration of a bearing defect diagnosis system according to an embodiment of the present invention. 1, the bearing defect diagnosis system of the present invention comprises a sensor unit 10, a signal processing unit 20, and a display unit 30.

Here, the sensor unit 10 is for detecting a vibration signal of the bearing, and is composed of a vibration detection sensor 11 installed in a bearing housing of the bearing. The vibration detection sensor 11 detects vibrations occurring in the bearing in real time.

The signal processing unit 20 uses the vibration signal of the bearing sensed by the sensor unit 10 to determine whether the bearing is defective or not. That is, the vibration signal of the bearing is filtered, and predetermined data is extracted by using the vibration signal values before and after filtering to determine whether or not the defect is defective. As shown in the figure, the signal processing unit 20 includes a sensor interface 21 for acquiring a vibration signal measured by the vibration sensor 11, and an apparatus for filtering and processing the obtained vibration signal, An analog-to-digital (A / D) converter 23, a temporary data storage device 24, a digital filter 25 and a signal processing device 26 are provided.

The display unit 30 includes a display device 31, an alarm device 32 and a data storage device 33 for displaying whether or not the bay ring is defective, informing an alarm, and storing data.

The vibration signal obtained from the sensor unit 10 is passed through the analog filter 22 and the digital filter 25 and converted into an energy concentration ratio Er and a complexity C of the vibration signal, As shown in FIG. The signal processing device 26 processes the vibration signal before and after filtering to extract the energy concentration Er and the complexity C of the signal and extracts the monitoring factor Sb based on the extracted data Er and C, To determine whether the bearing is defective based on the monitoring factor. The result of the determination as to whether or not the bearing is defective is displayed through the display unit 31 of the display unit 30. If there is a defect, the alarm condition is informed via the alarm unit 32, ). Such a series of operations may be performed by a control unit set in a predetermined programming mode. Hereinafter, the process of diagnosing a bearing defect will be described in detail.

2 is a flowchart illustrating a process for diagnosing bearing defects according to an embodiment of the present invention. As shown in the figure, the bearing defect diagnosis of the present invention measures a vibration signal generated in a bearing (S11), processes the measured vibration signal according to a predetermined algorithm (S12 to S14), determines whether the bearing is defective (S15 to S16), and a process of taking subsequent actions according to the result (S17). Hereinafter, each of the above processes will be described in detail.

1. Vibration signal measurement step ( S11 )

Rotating equipments generate vibration regardless of the state. In steady state bearings, there is no constant periodicity. Generally, it shows the characteristic of white noise and shows a very low flat spectrum in all frequency bands. On the other hand, in the case of a defective bearing, the periodicity associated with the bearing defect is shown, and in the frequency spectrum too much energy is concentrated in a specific frequency band. This specific frequency band is defined as the frequency corresponding to the natural frequency of the bearing.

[Formula 1]

Where fn is the natural frequency of the bearing, k is the stiffness of the bearing, and m is the mass of the bearing. Therefore, in the measuring step, a measurement can be made in a sufficiently wide frequency range including the natural frequency of the bearing.

2. Vibration signal processing step ( S12 ~ S14 )

The present invention defines a monitoring factor for determining the presence or absence of a bearing from a bearing vibration signal and determines the state of the bearing according to the value of the monitoring factor. In the process of defining the monitoring factor, the energy concentration (Er) obtained from the frequency spectrum and the complexity (C) of the vibration signal are used. Two parameters are used based on the following characteristics.

The vibration signal of the steady state bearing is mainly composed of random noise having no periodicity, and energy is not concentrated in a specific frequency band, and an even energy distribution is shown over the entire frequency band (refer to FIG. 3). That is, a large number of frequency components exist. Therefore, the complexity of a signal having a large value increases greatly as the frequency component increases. However, the vibration signal of the defective bearing generates a transient component or impulse periodically in accordance with the rotation of the equipment, and the intensity of the signal increases, so that the complexity of the signal decreases.

Accordingly, in the present invention, the vibration signal is filtered and the energy concentration Er and the complexity C of the signal are extracted based on the vibration signals before and after the filtering (x (t), x '(t) Lt; RTI ID = 0.0 > Sb < / RTI >

end. Filtering S12 )

The vibration signal (x (t)) of the measured bearing is extracted through the filtering process to extract only the signal (x '(t)) in the natural frequency band of the bearing. This is done for the purpose of eliminating the influence of the signal components irrespective of the state of the bearings to the maximum.

I. Energy concentration extraction ( S13 )

The energy concentration (Er) is calculated to determine the state of the bearing using the vibration signal x '(t) that has undergone the filtering process and the vibration signal x (t) before the filtering process. Energy concentration is defined as follows.

[Formula 2]

Here, X (f) and X '(f) represent the Fourier transform frequency spectrum (FFT) of the bearing vibration signal before and after filtering. That is, the energy concentration (Er) represents the ratio (sum of 1.0) of the energy sum included in the resonance frequency band and the entire frequency band of the bearing of interest. Since the vibration signal of the steady state bearing exhibits a uniform energy distribution in the whole frequency band, the energy concentration (Er) according to the state of the bearing is large because the energy of the vibration of the bearing having the defect is distributed in the natural frequency band of the bearing. . This difference is caused by the periodic transient component (impulse signal) caused by the bearing defects.

However, in general, vibration signals are mixed with various frequency components in addition to noise. Therefore, even though the filtering process is performed, noise existing in the pass band of the filter and other unnecessary frequency components exist, and it is incomplete to determine the state of the bearings only by the energy concentration Er. Therefore, in the present invention, the state of the bearing is determined using the complexity C of the signal as a parameter together with the energy concentration Er.

All. Complexity Extraction of Signal ( S13 ' )

The complexity of the signal is calculated by applying Lempel-Ziv Complexity. Lempel-Ziv Complexity is expressed as a value between 0 and 1 by non-dimensionalizing the occurrence frequency of a new pattern according to the repeated pattern of the signal. In the case of a very large number of frequency components such as random noise, the complexity increases to have a complexity close to "1 ". In the case of a signal having a periodicity such as a sine function, the complexity close to" 0 " ).

Therefore, if the bearing is defective, as the bearing rotates, the periodic transient component increases greatly due to the defect, resulting in a reduction of the signal complexity.

la. Monitoring factor  Calculation( S14 )

The monitoring factor Sb is calculated as follows using the energy concentration (Er) obtained previously and the complexity (C) of the signal.

[Formula 3]

Here, the monitoring factor Sb has a value of 0 to 1. This is because both energy concentration (Er) and signal complexity (C) are between 0 and 1. The presence or absence of defects in the bearing through the monitoring factor (Sb) was defined based on the characteristics shown in [Table 1]. When the monitoring factor Sb is calculated according to the characteristics shown in Table 1, it has a small value (i.e., a value close to "0") in a steady state, Quot; 1 ").

Bearing condition Energy concentration (Er)
(Maximum value = 1.0) The complexity of the signal (C)
(Maximum value = 1.0) The monitoring factor (Sb)
(Maximum value = 1.0) Steady state that The that Defect Status The medium medium

The bearing vibration signal in the fault state is composed of a combination of several transient components having a periodicity, and generally has a value equal to or greater than the complexity (0.2 to 0.3) of the sinusoidal wave.

3. Judgment and follow-up steps ( S15 ~ S17 )

One of the methods for determining the state of the bearing using the monitoring factor Sb defined above can be determined based on the degree of change of the monitoring factor Sb with time. However, there is a need for a reference value that can be determined for more efficient and stable determination. In the present invention, the reference value is presented based on the vibration signal of the steady state bearing as follows.

The vibration signal of the steady-state bearing has a value which is opposite to that of the energy concentration (Er) and the complexity (C) of the signal, which are two parameters as shown in Table 1 above. In other words, taking the ideal steady state bearing as an example, the energy concentration Er has a value near "0", but the signal complexity C has a value of "1". Here, the vibration signal of the ideal steady state bearing means that it is composed only of random noise. Therefore, the monitoring factor Sb has a value near "0 ".

However, in the actual case, it is difficult to have a value of the monitoring factor (Sb) close to exactly "0" even in the vibration signal of the steady state bearing because the nonlinearity existing in the equipment itself and the background noise Background noise is present. In consideration of this point, a certain margin is given as a reference value for determining the bearing condition. In the present invention, a value between 0.1 and 0.2 is presented as the reference value (Ref) of the monitoring factor. However, the value is a value that can be adjusted by the user depending on the characteristics and environment of the monitored facility.

If it is determined that there is a defect in the bearing (that is, when the monitoring factor value is larger than the reference value) by the judgment formula comparing the calculated monitoring factor Sb and the reference value Ref, an alarm is generated, A maintenance plan for the treatment is established. If it is determined that there is no defect in the bearing, the vibration signal is repeatedly measured, and the state of the bearing is continuously monitored through the signal processing step shown in the present invention.

Experimental Example

The performance of the steady state bearing and the defective bearing is verified by the above process. FIG. 3 is a graph showing a frequency spectrum of a steady state in the bearing defect diagnosis system of the present invention, and FIGS. 4 and 5 are graphs showing frequency spectra of the inner and outer ring defect states in the bearing defect diagnosis system of the present invention.

First, as shown in FIG. 3, the frequency spectrum of the steady state bearing vibration signal shows an even energy distribution over the entire frequency band in which the measurement is made. However, in the frequency band of 4000Hz or more, high-frequency components due to peripheral vibration and non-linearity of bearings are partially appeared. In other words, it can be seen that the oscillation of the actual steady state bearing is partially different from the frequency spectrum when the ideal random noise is present. Therefore, several parameters must be considered to determine the condition of the bearing.

On the other hand, as shown in FIGS. 4 and 5, it can be seen that the frequency spectrum of the vibration signal of the bearing having the inner ring defect and the outer ring defect has a large amount of energy gathered from 2000 Hz or more. Also, in the case of the bearings shown in Figs. 3 to 5, it is shown that the natural frequency of the bearing is located between 2000 Hz and 4000 Hz.

3 to 5, a filter is applied to a frequency band between 2000 Hz and 4000 Hz according to the bearing defect diagnosis method of the present invention to determine the energy concentration Er and the signal complexity C ) And the monitoring factor (Sb) is obtained as shown in [Table 2]. In the example of [Table 2], 0.15 was used as the reference value Ref for judgment. As shown in [Table 2], it can be seen that the value of the monitoring factor (Sb) shows a clear difference depending on whether or not the bearing is defective.

Bearing condition Energy concentration (Er) The complexity of the signal (C) The monitoring factor (Sb) Judgment result
(Standard = 0.15) Steady state (Figure 3) 0.11 0.82 0.09 normal The inner ring defect (Figure 4) 0.50 0.60 0.30 flaw Outer ring defects (Figure 5) 0.68 0.44 0.30 flaw

In order to test the conformity of the present invention, the results of judging the state of the bearings by using vibration signals for seven cases are summarized in Table 3.

No. Bearing condition Energy concentration (Er) The complexity of the signal (C) The monitoring factor (Sb) Judgment result
(Standard = 0.15) One normal 0.089 0.820 0.07 normal 2 Outer ring defect 1 0.709 0.459 0.33 flaw 3 Outer ring defect 1 0.443 0.742 0.33 flaw 4 Outer ring defect 3 0.598 0.381 0.23 flaw 5 Inner ring defect 1 0.486 0.693 0.34 flaw 6 Inner Wheel Defect 2 0.576 0.527 0.30 flaw 7 Inner Wheel Defect 3 0.688 0.459 0.32 flaw

As shown in Table 3, it can be seen that the monitoring factor (Sb) defined in the present invention for all cases accurately determines the state of the bearing.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments.

10: Sensor unit
20: Signal processor
30:
x (t): Vibration signal measured by the sensor
x '(t): The vibration signal passed through the filter
X (f): Fourier transform frequency spectrum of the vibration signal (x (t)) measured by the sensor
X '(f): Fourier transform frequency spectrum of the vibration signal (x' (t)
Er: Energy concentration
C: complexity of signal
Sb: Bearing condition monitoring factor
Ref: Bearing condition judgment reference value

Claims (6)

A system for diagnosing defects in bearings of rotating equipment,
A sensor unit for measuring a vibration signal of the bearing;
A filter for extracting a natural frequency band signal with respect to the measured vibration signal and a frequency spectrum energy sum of a frequency band of interest and an entire frequency band from a vibration signal before filtering and a vibration signal before filtering, The complexity C of the signal defined by the complexity calculated from the energy concentration Er and the interest band vibration signal is calculated and the monitoring factor value Sb is extracted from the energy concentration Er and the complexity C of the signal, And a signal processing unit for determining whether the bearing is defective by comparing the extracted monitoring factor value (Sb) with a predetermined reference value (Ref). And
And a display unit including a display device for indicating presence or absence of a defect in the bearing, and an alarm device for generating an alarm when a defect is generated in the bearing. A method for diagnosing a defect occurring in a bearing of a rotary facility,
(a) a vibration signal measurement step of measuring a vibration signal of a bearing at a sensor part provided in a housing of a bearing;
(b) acquiring and filtering the vibration signal measured in the step (a), calculating an energy concentration (Er) defined by a ratio of the frequency spectrum energy sum of the interest band and the entire frequency band, A signal processing step of obtaining a complexity C of a signal defined by the signal Sb and calculating a monitoring factor Sb therefrom; And
(c) comparing the monitoring factor (Sb) of the step (b) with a reference value (Ref) to determine a state of the bearing. delete 3. The method of claim 2,
Wherein the monitoring factor Sb is a parameter defined as a dimensionless value between 0 and 1 from the energy concentration Er and the complexity C of the signal. 5. The method of claim 4, wherein step (c)
Wherein the reference value Ref is set to a value between 0.1 and 0.2, and when the monitoring factor Sb exceeds the reference value Ref, it is determined that there is a defect in the bearing. 3. The method of claim 2,
and (d) generating an alarm by displaying a result of the determination in step (c) if a defect is present in the bearing.

Images