KR101406778B1 - Method for monitoring machinery health based on fictitious frequency response function and system using the same - Google Patents

Abstract

A method for monitoring machinery health based on a fictitious frequency response function is provided. The method for monitoring machinery health comprises the steps of measuring a first vibration signal; measuring a second vibration signal different from the first vibration signal; calculating a fictitious frequency response function (fictitious FRF) of the first and second vibration signals; extracting a group delay from the calculated fictitious FRF; and analyzing the fictitious FRF and the group delay. As a result, the method according to the present invention can detect a subtle defect of a monitored object early by monitoring a spectrum of magnitudes and a spectrum of phases separately.

Description

TECHNICAL FIELD [0001] The present invention relates to a method for diagnosing an anomaly of a machine based on a virtual frequency response function, and a diagnosis system using the virtual frequency response function,

The present invention relates to a method for diagnosing a mechanical abnormality and a diagnostic system using the same, and in particular, by using a virtual frequency response function (virtual FRF), the sensor can not be attached to a desired position due to environmental structural constraints, The present invention relates to a diagnosis method and a diagnosing system for a malfunctioning machine based on a virtual frequency response function capable of realizing an early diagnosis even when health monitoring methods can not be effectively used.

Generally, mechanical diagnosis methods using vibration signals are performed by measuring vibration signals by attaching sensors (mainly acceleration sensors) adjacent to major parts such as bearings and gears, and then measuring RMS (or standard deviation) of vibration signals, Kurtosis ), Power spectrum or envelope spectrum, etc., to determine whether or not the machine is abnormal. Especially, frequency analysis method using power spectrum is most used.

Conventional methods simultaneously measure signals of a plurality of channels using a plurality of sensors, but the method of analyzing each channel is the same and basically belongs to a single channel data analysis technique. These methods can vary sensitively depending on the location where the sensor is attached, and in some cases due to environmental constraints, such as temperature, humidity, electrical and magnetic factors, and structural constraints, It is often the case that it can not be attached. In this case, the reliability of the diagnosis result may be greatly deteriorated.

Such a conventional mechanical diagnostic method is shown in detail in FIGS. 1 to 3 are views for explaining a mechanical diagnostic method according to the prior art.

Referring to FIG. 1, in most rotating machines, an abnormal signal due to various defects is transmitted to other parts through the bearing part, and defects of bearings among many parts occur most frequently. Therefore, an acceleration sensor (A ₁ ) is attached to the bearing housing or a region adjacent to the bearing housing to measure the vibration signal, and various kinds of analysis methods such as RMS or standard deviation are used to identify the presence or absence of the defect.

에 의해 계산된다.RMS or standard deviation is a measure of the overall size of the signal, and the simplest way than the oscillation signals are used under a greater vibration signal at the time when the fault assumptions of normal use, x of a discrete signal measured by the sensor ( n)

Lt; / RTI >

와 같이 계산될 수 있다. 이 경우,

이다.Referring to FIG. 2, although the RMS represents the characteristics of the entire vibration signal in the time domain as one index, many vibration signals exhibit unique characteristics depending on the frequency. Especially, in the case of a defect, specific frequencies tend to increase or decrease have. Therefore, the power spectrum is used when the vibration signal is decomposed and analyzed by frequency in the frequency domain. Physically, it has the meaning of dividing the dispersion (or mean square) value of the vibration signal by each frequency. Let X _i (t) be the measured vibration signal for each section in the time domain, and let X _i ( f ) be the Fourier transformed value.

Can be calculated as follows. in this case,

to be.

These conventional methods can be summarized in the diagram of FIG. 3, which is considered to analyze only the output signal in a single input ( s ) and a single output (x) system. Referring to FIG. 3, s (t) denotes a vibration component due to mismatch, unbalance, or bearing defect, and x (t) denotes a signal measured by the acceleration sensor A ₁ .

The system is a transfer function corresponding to the path between the vibrating point of the excitation point and the vibration point of the response point. Since the excitation component s (t) corresponding to the input is generally impossible to measure, only the output component x Defect analysis is performed by signal processing methods. Therefore, if the measuring sensor is attached to a point as close to the point as possible, it can detect the abnormality in the early stage. If it is attached to a distant place due to environmental or structural problems, the abnormality signal included in the output component x (t) Which is difficult to detect.

An object of the present invention is to solve the above-mentioned problems, and it is an object of the present invention to provide a fault diagnosis method based on a virtual frequency response function capable of early detection of a minute defect of an object by independently observing a spectrum with respect to size and phase And a diagnostic system using the same.

It is another object of the present invention to analyze the actual shape of complex data by using a group delay to represent a phase change with respect to a virtual frequency response function in the case of a defect and in the case of a defect, And a diagnostic system using the virtual fault response function based on the virtual frequency response function.

In order to achieve the above object, a method for diagnosing an anomaly of a machine based on a virtual frequency response function according to an embodiment of the present invention includes the steps of: i) measuring a first vibration signal; ii) (Iii) calculating a virtual frequency response function (virtual FRF) of the first vibration signal and the second vibration signal, iv) extracting a group delay from the calculated virtual FRF, and iv) Analyzing the virtual FRF and the group delay.

In addition, according to the method for diagnosing an anomaly of a machine based on the virtual frequency response function according to the embodiment of the present invention, the step (i) includes the steps of obtaining a steady state signal at a plurality of different measurement points including a first measurement point and a second measurement point And calculating a first cross spectrum from the data obtained from the plurality of different measurement points.

According to the method for diagnosing an anomaly of a machine based on the virtual frequency response function according to the embodiment of the present invention, the step (ii) is a step of detecting a defect state signal at a plurality of different measurement points including the first measurement point and the second measurement point And calculating a second cross spectrum from the data obtained from the plurality of different measurement points.

According to an embodiment of the present invention, the virtual FRF includes a first mutual spectrum for the first vibration signal and a second mutual spectrum for the second vibration signal, Is calculated.

In addition, according to the method for diagnosing anomalous machine based on the virtual frequency response function according to the embodiment of the present invention, the virtual FRF is defined by the following formula (2).

&Quot; (2) "

Also, according to the method for diagnosing anomalous machine based on the virtual frequency response function according to the embodiment of the present invention, the group delay of the step iv) extracts the phase of the virtual FRF and indicates a phase change from the extracted phase .

According to the method for diagnosing an anomaly of a machine based on a virtual frequency response function according to an embodiment of the present invention, the group delay is defined by Equation (3).

&Quot; (3) "

In addition, according to the method for diagnosing an anomaly of a machine based on a virtual frequency response function according to an embodiment of the present invention, the step (v) is a step of analyzing whether a mechanical defect and an anomaly are diagnosed from an analysis of a vibration signal.

According to another aspect of the present invention, the first vibration signal and the second vibration signal are measured using an acceleration sensor.

As described above, according to the fault diagnosis method based on the virtual frequency response function according to the embodiment of the present invention and the diagnosis system using the same, it is possible to observe the spectrum of the magnitude and the phase independently, Can be obtained.

In addition, according to the fault diagnosis method based on the virtual frequency response function and the diagnosis system using the virtual frequency response function according to the embodiment of the present invention, a group for representing the phase change of the virtual frequency response function when there is no fault and when there is a fault By using a group delay, an effect of being able to express how large the phase change occurs in a specific frequency band is also obtained by analyzing the actual shape of the complex data.

1 to 3 are views for explaining a mechanical diagnostic method according to the prior art.
FIGS. 4 to 6 are diagrams for explaining a mechanical abnormality diagnosis method based on a virtual frequency response function according to an embodiment of the present invention.
FIG. 7 is a flowchart for explaining a malfunction diagnosis method based on a virtual frequency response function according to an embodiment of the present invention.
8 is a diagram for explaining an experimental example of a mechanical abnormality diagnosis method according to an embodiment of the present invention.
9 to 11 are graphs for explaining experimental results of the mechanical abnormality diagnosis method according to the embodiment of the present invention.
FIGS. 12 and 13 are diagrams for explaining the diagnostic method according to the prior art in comparison with experimental results of the present invention.

Hereinafter, the present invention will be described in detail with reference to the embodiments shown in the accompanying drawings, but the present invention is not limited to the illustrated embodiments.

These and other objects and novel features of the present invention will become more apparent from the description of the present specification and the accompanying drawings.

Hereinafter, a fault diagnosis method based on a virtual frequency response function according to an embodiment of the present invention and a diagnosis system using the same will be described in detail with reference to FIG. 4 to FIG.

FIGS. 4 to 6 are diagrams for explaining a mechanical abnormality diagnosis method based on a virtual frequency response function according to an embodiment of the present invention. FIG. 7 is a flow chart for explaining a malfunction diagnosis method based on a virtual frequency response function according to an embodiment of the present invention, and FIG. 8 is a view for explaining an experimental example of a malfunction diagnosis method according to an embodiment of the present invention . FIGS. 9 to 11 are graphs for explaining experimental results of the mechanical anomaly diagnosis method according to the embodiment of the present invention, and FIGS. 12 and 13 illustrate the diagnostic method according to the prior art, Respectively.

4, a fault diagnosis system based on a virtual frequency response function according to an embodiment of the present invention includes a plurality of vibration sensors mounted at arbitrary different positions S1 and S2 of an object to measure vibration signals, do. The vibration sensor may be an acceleration sensor, but it is not limited thereto and other sensors for measuring the vibration of the object may be used. In this case, the object includes a motor, a bearing, a gear, and a rotary shaft, and includes all components connected to a vibration source or a vibration source to generate vibration different from the vibration source.

The data measured from the vibration sensor include both a first vibration signal without a mechanical defect and a second vibration signal caused by a mechanical defect. That is, the first vibration signal means a signal in a steady state during driving of the mechanical device, and the second vibration signal means a signal in an abnormal state occurring during driving of the mechanical device. For example, the second vibration signal includes all of the mismatch, imbalance, or bearing defects of the mechanical device, and may be the same signal as the steady state signal when the mechanical device is normally driven.

The first vibration signal may be measured by mounting a plurality of acceleration sensors at different positions in a mechanical device, and the second vibration signal may be measured at the same position as the first vibration signal. The first and second vibration signals are stored in a data storage unit provided in the system, and the mutual spectrum calculator calculates a first mutual spectrum for the first vibration signal and a second mutual spectrum for the second vibration signal using the data, Calculate the mutual spectrum.

The first mutual spectrum and the second mutual spectrum may be calculated using a conventional computer program. The virtual FRF is calculated using the first and second inter-spectra, from which a group delay representing a phase change with respect to magnitude, phase, and frequency of the virtual frequency response function, Is extracted by the element extracting unit. Therefore, the system determines whether the mechanical device is defective or abnormal based on the size, phase, and group delay of the virtual FRF extracted from the element extracting unit.

)이 계산된다.Referring to FIGS. 5 through 7, the method for diagnosing an anomaly of a machine according to the embodiment of the present invention includes measuring a first vibration signal (S71). The first vibration signal includes a steady-state vibration signal free from mechanical faults, and includes a first measurement point S1 and a second measurement point S2 (for example, ) From a plurality of different measurement points. The first vibration signal may be measured using an acceleration sensor. If there is no defect, respectively, the measured signal from the first measurement point (S1) and the second measuring point (S2) is defined as ₁ U (t) and U ₂ (t). The first mutual spectrum corresponding to the first vibration signal after the first vibration signal is measured

) Is calculated.

는 매우 작게 나타난다. 즉,

(상기 제1상호 스펙트럼 값이 정확하게 0의 값을 가지지는 않음).Since the first vibration signal has no defect, there is no particularly noticeable excitation element in the frequency band of interest, and if there is a sufficient distance between the first measurement point S1 and the second measurement point S2, Can be regarded as almost nonexistent. Therefore, the first mutual spectrum < RTI ID = 0.0 >

Is very small. In other words,

(The first mutual spectral value does not have a value of exactly zero).

)이 계산된다. 상기 제2상호 스펙트럼은 컴퓨터 프로그램을 이용하여 제1상호 스펙트럼과 동시 또는 시간을 달리하여 계산될 수 있다.After the step S71, the second vibration signal is measured by mounting the sensor at the first measurement point S1 and the second measurement point S2 (S72). The second vibration signal includes a vibration signal in the abnormal state where there is a mechanical fault, that is, a new vibration element s (t) occurs. If there is no mechanical fault (no new vibration element s (t) ) May be the same signal as the first vibration signal. From the second vibration signal to a second mutual spectrum (

) Is calculated. The second mutual spectrum may be computed simultaneously or in time with the first mutual spectrum using a computer program.

If a new vibrating element s (t) such as a bearing defect or a mismatch occurs during operation in a steady state without a defect, the vibrations will propagate through the first path (path 1; H1) from the excitation point to the first measurement point S1 And reaches along a second path (path 2; H2) different from the first path up to the second measurement point S2. Each x ₁ a direct response of the new with elements s (t) due to the defect (t) and x ₂ of the signal measured at each measurement point defined, and a (t) x _m1 (t) and x _m2 (t) . Thus, the entire system can be represented by the diagram of FIG. In this case, the mutual spectrum between the measured signals x _m1 (t) and x _m2 (t) is expressed by the following equation (1) because the correlation is large only between the direct responses x ₁ (t) and x ₂ Can be displayed.

및 새로운 결함이 발생한 경우의 제2상호 스펙트럼

을 이용하여 아래 [수학식 2]와 같이 구성될 수 있다.After the step S72, a virtual frequency response function (virtual FRF; H _fic (f)) of the first vibration signal and the second vibration signal is calculated (S73). The virtual frequency response function H _fic (f) is a function of the first mutual spectrum

And a second mutual spectrum when a new defect occurs

Can be expressed by the following equation (2).

는 매우 작은 값이기 때문에 결함이 있는 경우 가상 FRF의 크기 |H_fic(f)|는 매우 크게 증폭되어 나타나게 되어 미세한 결함도 검출이 가능하다. 따라서, 본 발명은 피측정물의 조기 이상 진단에 매우 효과적이다. 또한, 증폭되어 나타나는 성분은 위상에도 반드시 반영되기 때문에 해당 주파수 성분의 위상 φ_fic(f)을 동시에 관찰하면 보다 명확한 결함 정보를 알 수 있다.In Equation (2), it can be seen that the defective component due to the new excitation element appears well in the virtual FRF. The first mutual spectrum in the absence of defects

Is very small. Therefore, if there is a defect, the magnitude of the virtual FRF | H _fic (f) | is amplified so that a minute defect can be detected. Therefore, the present invention is very effective for early diagnosis of the object to be measured. In addition, since the components appearing in the amplified state are also reflected in the phase, more accurate defect information can be obtained by observing the phase _fic (f) of the corresponding frequency component at the same time.

If the phase φ _fic (f) is directly compared after the step S73, the analysis may be difficult. The group delay is extracted from the calculated virtual FRF to analyze whether the fault is a mechanical defect or not (S74 and S75). The group delay can indicate how large the phase change occurs in a particular frequency band and is defined by Equation (3) below.

을 구하고, 결함이 있다고 판단되거나, 정기적으로 동일한 지점에서 진동신호를 측정하여 두 신호 사이의 제2상호 스펙트럼

를 구한다. 상기 제2상호 스펙트럼을 구한 후에, 제1상호 스펙트럼 및 제2상호 스펙트럼 사이의 관계를 나타내는 가상 FRF, 즉,

를 구한 후에, 가상 FRF의 크기 |H_fic(f)|, 위상 φ_fic(f) 또는 그룹지연 τ(f)를 분석하여 이상 주파수 성분이 나타나는지 관찰하여 결함 유무를 판단한다.
Therefore, the present invention first measures vibration signals at two points at a sufficient distance in a steady state to determine a first mutual spectrum

And it is judged that there is a defect or the vibration signal is periodically measured at the same point so that the second mutual spectrum

. After obtaining the second mutual spectrum, a virtual FRF representing the relationship between the first mutual spectrum and the second mutual spectrum,

After the calculated size of the virtual FRF | H _fic (f) | observed appear over a frequency component by analyzing the phase φ _fic (f) or the group delay τ (f) is determined for defects.

<Experimental Example>

Referring to FIG. 8, after making an artificial defect in the outer ring of the ball bearing and the ball itself, the vibration was measured at two measurement points (S1, S2) to compare the power spectrum according to the prior art with the present invention. In this case, the rotor is rotating at 1800 rpm, the bearing outer ring defect frequency is about 108 Hz component, and the defect frequency of the ball itself is about 78 Hz component.

9 to 11, the present invention is characterized in that the bearing outer ring defect frequency (108 Hz) component and the second harmonic component (216 Hz) are remarkably remarkable in the magnitude and phase spectrum, and the ball self- Hz) are also amplified, but the other components are not well visible, so the discrimination power is very good. In addition, the phase change occurs largely around 108 Hz and 216 Hz, but the 78 Hz component is not visible. However, observing the group delay shows a large change in all three components.

Referring to FIGS. 12 and 13, in the power spectrum analysis according to the related art, high-frequency components in which a rotational frequency (30 Hz) component and a plurality of excitation components are combined are prominent in the spectrum at the measurement point S1, The power supply frequency (180 Hz) component of the power supply is also shown to some extent. And the spectrum at the measuring point S2 is very close to the motor, so the power frequency (180 Hz) component of the three-phase induction motor is very large and the other components are not noticeable. In both cases, the fault frequency components are not well represented because the measuring point is far from the bearing part.

Although the present invention has been described in detail with reference to the above embodiments, it is needless to say that the present invention is not limited to the above-described embodiments, and various modifications may be made without departing from the spirit of the present invention.

Claims (10)

delete I) measuring a first vibration signal;
Ii) measuring a second vibration signal different from the first vibration signal;
Iii) calculating a virtual frequency response function (virtual FRF) of the first vibration signal and the second vibration signal;
Iv) extracting a group delay from the calculated virtual FRF; And
And v) analyzing the virtual FRF and the group delay,
The step i)
Measuring a steady state signal at a plurality of different measurement points including a first measurement point and a second measurement point; And
And calculating a first cross spectrum from data obtained from the plurality of different measurement points. &Lt; Desc / Clms Page number 19 &gt; I) measuring a first vibration signal;
Ii) measuring a second vibration signal different from the first vibration signal;
Iii) calculating a virtual frequency response function (virtual FRF) of the first vibration signal and the second vibration signal;
Iv) extracting a group delay from the calculated virtual FRF; And
And v) analyzing the virtual FRF and the group delay,
The step ii)
Measuring a signal of a fault condition at a plurality of different measurement points including a first measurement point and a second measurement point; And
And calculating a second cross spectrum from the data obtained from the plurality of different measurement points. &Lt; Desc / Clms Page number 19 &gt; I) measuring a first vibration signal;
Ii) measuring a second vibration signal different from the first vibration signal;
Iii) calculating a virtual frequency response function (virtual FRF) of the first vibration signal and the second vibration signal;
Iv) extracting a group delay from the calculated virtual FRF; And
And v) analyzing the virtual FRF and the group delay,
Wherein the virtual FRF is calculated from a first mutual spectrum for the first vibration signal and a second mutual spectrum for the second vibration signal. 5. The method of claim 4,
Wherein the virtual FRF is defined by the following equation (2): &quot; (2) &quot;
&Quot; (2) &quot;

6. The method according to any one of claims 2 to 5,
Wherein the group delay of step iv) extracts a phase of the virtual FRF and represents a phase change from the extracted phase. The method according to claim 6,
Wherein the group delay is defined by: &lt; EMI ID = 3.0 &gt;
&Quot; (3) &quot;

delete delete A machine abnormality diagnosis system based on a virtual frequency response function, characterized in that a mechanical defect and an abnormality are diagnosed by using any one of the methods selected from the items 2 to 5.

Images