JPS59230492A - Monitoring method of rotary electric machine - Google Patents

Abstract

PURPOSE:To enable to detect the malfunction of a rotary electric machine in an early period by comparing the sampled pattern obtained by the n-dimensional vector produced by Fourier analysis of the detection signal with the reference pattern and with the vibrating amplitude of the signal, and discriminating the malfunction state. CONSTITUTION:After the detection signal of a rotary electric machine 1 is analyzed by a Fourier analyzer 6 to calculate a power spectral density or coherence, the sampled pattern of n-dimensional vector is obtained in a sampled pattern collecting unit 6, the frequency value of the sampled pattern is standardized by the rotary frequency, and further standardized by the dispersion of the signal. A comparator 8 compares the distance between the corrected sampled pattern and the reference pattern with the amplitude of the vibration of the rotating frequency in RMS value to detect the malfunction. In this manner, the reliability of the detected result of the malfunction state can be improved.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to a nuclear power plant whose rotation If
A Concerning a monitoring method for rotating equipment that detects equipment abnormalities at an early stage and prevents their failure [Technical background of the invention and its problems] For example, large liquid metal cooling type For fast breeder reactor power plants, an operating method is planned in which the reactor output/flow rate ratio is maintained constant by manipulating control rods and coolant flow rates. In addition, in the operation of boiling water reactors currently in operation, changes such as increasing or decreasing the reactor output are made by controlling both the control rods and the core flow rate. In order to do this, a method of controlling using control equipment, particularly rotating equipment such as a pump or an electric motor, is adopted. Therefore, if a signal-1 abnormality occurs in these rotating equipment, it may have a double negative impact on the nuclear power plant (nuclear power plant), and it is expected to have an extremely large impact on its operating rate. The rotation of electric motors, etc. (equipment) is constantly monitored, and if any abnormality occurs in these rotating equipment, the abnormality is detected at an early stage to prevent failure of the rotating equipment and prevent nuclear power plants. It is essential to operate the C2 in a healthy manner and to increase its operating rate.

The following prior art has been proposed as a method for constantly monitoring rotating equipment such as pumps or electric motors using such contacts.

Pump or frost: Rotating equipment such as motors, or drive shafts of these rotating equipment (- mounted displacement gauges, speedometers.

The signal obtained from the accelerometer is subjected to Fourier analysis to express the sample pattern by a linear vector If the reference pattern in is expressed by a vector μ having the same configuration, the distance is the vector X.

Using the standard deviation σL of the elements xl of μ, μ; and the element X in the normal state, and the standard deviation vL of the element μ, the following (1)
Defined by Eq.

D= )i, 1. z:4-μb1'-(xL0μ
m)"/(aal'+bvb') -------(1
)-1 Here, D is y, ==l, city distance s if r-s=0, Mahalanob if t=r=1+'=2
This is called the distance of ls, and a and b are constants determined empirically.

In equation (1) above, if the distance is greater than a predetermined threshold value α, it is determined that the rotating equipment such as the pump or the motor is in an abnormal state ζ.

However, it is known that the vibration of a rotating device such as a pump or an electric motor generally changes in amplitude due to a change in the rotational frequency fo.
Since the amplitude of the frequency component proportional to is large, the second appears as a peak on the frequency pattern.

Therefore, even in normal conditions, the current rotational frequency fx
However, even if the rotational frequency jp changes slightly from the time when the reference value μ was learned, the "distance" D shows a large value, and the rotating equipment such as the pump or electric motor is determined to be in an abnormal state. There is a risk of

[Object of the Invention] The present invention was developed in view of the circumstances of the sr3fj article.

The purpose of this is to determine whether the change in vibration amplitude is normal or not, and to detect abnormalities in rotating equipment at an early stage. There are two C types that provide a method for monitoring equipment.

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention consists of a power spectrum density or coherence obtained for each frequency point by sequentially (Ni-Fourier analysis) a detection signal from a rotating device. In order to compare the sample value and its reference value in a normal state, the frequency value is normalized by the rotation frequency of the signal. The sample pattern based on the sample value C2 is compared with the reference pattern based on the reference value, and the vibration amplitude is further compared using the effective value of the amplitude of the signal, thereby determining an abnormality in the rotating equipment. Monitoring method 1 for rotating equipment: When the distance between the sample value and the reference value is greater than a predetermined threshold, it is determined that the signal pattern has changed and the rotating plate equipment is abnormal. In addition, it is displayed that the rotating equipment is in an abnormal state due to an abnormality C2 in the effective value of the amplitude of the signal.

Next, the principle of the present invention will be explained in detail.

Now, the signal #(t) detected from a rotating device such as a pump or an electric motor is converted into a digital signal G(L・Δt), L=i~2Nl at a sampling interval Δt, and then the analysis is performed. The reference value of the power spectral density Y(f) obtained for each frequency equinox based on Yμ.

Assume that the sample value is Yx, and that the reference value Y/J and the sample values Y, z- have a relationship as shown in FIG. Then, if the frequency resolution of this analysis is △f, and the analysis frequency is △f-N・Δf, the dispersion value s2 of the signal is lower G]
It is expressed by equation (2).

S2=/Fill Δj −−〜−−−−−−−−−−
(2) In ζ-ren, yi = Y<i・Δf).

Next, if we normalize the frequency 17X with the rotational frequency fo, we can see that the standard value Yμ and the sample value Y
, match the peak positions of . This can be done concretely as follows. That is, let p = △j ,, 7 u-mf
It is assumed that a reference vector μ is constructed from the logarithmically transformed value of the reference value Yμ of the power spectrum density WY(f) at .

At this time, the th element of the sample vector Xp: cL is
The logarithmically transformed value including y in the following equation (8) obtained by interpolating from the sample 51+'/J+□ of J·Δf and the power spectrum density Y at (241)Δfr2) may be used.

J/A = yi + (J'-J) ・”
/i+1-”/j) −−−−−−CB+here,!
= 1' integer part 1 ``''/L・ta/jp In this way, i2 and the sample value Y5. The corrected reference value Yp is y, p+
Let it be YPp.

In addition, when Y is the power spectrum intelligence, as shown in FIG. 3, Ypp/Ypp/ From the logarithmically transformed values of 8X and Y p/Ss,
S and sample vector X8, i.e. μl]:=−μ,
-jn(Sp'') -Yamaichi---------(4
)Xs = Xp −% (%”) −−−−−−
------------1 (5). Please, JrLYI'
p''l'p + Ln Yzp =Xp.

Then, when abnormality detection is performed using power spectral density, C2 is s.When using D8° coherence, the "distance" between XS and μ8 is C'''-1et Ygp and Y/Ip.
The "distance JD" between the vectors Xp and μ, which are formed without performing logarithmic transformation (JM), is compared with the value α.

By doing this, as illustrated in FIG. 3, an abnormality is detected only when the shape of the pattern changes.

On the other hand, whether the change in vibration amplitude is a normal change caused by a change in rotation speed is determined by the effective value of the amplitude (Root
t-mean-square (hereinafter referred to as RM8fiQ)
8 or the dispersion value S2 as follows. That is, the initial formation of the motion of a nuclear power plant is carried out at the beginning of the trough operation cycle. From S ;F(f
) can be added to 0.

When learning the reference value Y of the power spectral density Y, the standard deviation W of the RMS value S of the amplitude is obtained at the same time.This index is a function of the rotation frequency f. Therefore, by setting w''=='V, it is assumed that an amplitude abnormality has occurred when the following equation (6) is satisfied from these relationships.

(8,-FgeJ) )” > C・V (lx)
------- (6) Here, C is a constant determined empirically.

[Embodiment of the Invention] An embodiment of the present invention will be described with reference to the block diagram of FIG.

First, a displacement meter attached to a rotating device 1 such as a pump or an electric motor or a drive shaft of such rotating device.

A signal f(, t) obtained from a speedometer, an accelerometer 1, a rotation speed needle, a rotation speed meter 2, etc. is sent to a sample hold circuit l of a signal input device 3 from a control command number from a control device 11. The analog input signal (i) is converted into a digital signal (L, Δt) by the A/D conversion circuit.
) and then stored in the data storage device 4I.

This digital signal G (L・△t) is transmitted to the data retention device 4
Then, the Fourier analyzer 5 calculates the high-speed vector density or coherence. In the next stage sample pattern correction device @6, the data storage device 4
In response to the input from the Fourier analyzer 5, a sample pattern X of a quadratic vector consisting of power spectral density YCf) or coherence C(f) at the rotational frequency is expressed by each element xL. On the other hand, in the reference data storage device 7, for Fourier analysis of the detection signal of rotating equipment in a normal state, there are various data representing the fundamental pattern μ of the fourth-order vector composed of the Jorhower spectral density Y(7) or the coherence C(f). The element μ is saved. In this sample pattern correction device 6, the frequency value of the sample pattern X is further normalized by the rotational frequency fo. Normalization Y/sZ is performed by the variance S2 of Yxp/sZ to obtain a corrected sample pattern Yxp/sZ. In the comparison device @ of the next stage, based on the command C2 from the control device 11, the corrected sample pattern Yxp/s, 2 and the reference C2 corrected and stored in the reference data storage device 7 in the same manner as above. Calculate the distance between pattern pp/Bp2. This distance calculation is performed using the formula (4) above, that is, μ6=μp
-1n (SP2) and the above formula (5), that is, Xs
= Xp - J, rL (5r2) It is performed from l2. When using the power spectral density as described above, the distance D between the reference vector μ and the sample vector Xs is
8, or if coherence is used, the vector μ, which is constructed as it is from the reference value Ypp and the sample value Yp, and Xp
Calculate the distance Dp from the

The respective distances Ds and Dp are then compared with respective predetermined thresholds. In this way, the distances Ds, D
By comparing the shapes of the sample pattern and the reference pattern, it is possible to compare the shapes of the sample pattern and the reference pattern. , that is, (s:IneeF'gework))2
〉C・■Gework) Performed from C2, the next stage output display device 9g
= is the RMS value S of this distance Ds, Dp and the vibration amplitude
The trend is shown, and whether the abnormality is in the shape of the vibration frequency pattern (vibration mode) or only in the amplitude is displayed. If an abnormality is found in the distance Ds, Dp or the RMS value SC, an alarm will be issued by the alarm generator 10.

Next, an embodiment of the rotating equipment monitoring method of the present invention explained above will be explained with reference to the flowchart of FIG.
-Explain.

First, when a monitor start command is issued, a detection signal p(f) from a rotating device such as a pump or an electric motor is read.
Next, this detection signal #(, t) is input to an A/D converter, and the analog signal t(f) is converted into a digital signal G(A・Δt).
≦2 converted and saved. The digital signal G(χ) is taken out periodically or at any time and subjected to fast Fourier transform (G)
Then, the power spectral density or coherence Y) is calculated. (Hereinafter, the case of power spectral density will be explained using two i.) Then, Konohauer spectral density [Y
The sample pattern
express. The frequency of this sample pattern X is 11! Normalization '/fo by the rotation frequency fol- of t is performed, and further normalization by the signal variance s2 )'/S2 is performed to obtain a corrected sample pattern 7 Yxp/B,2. This corrected sample pattern Yxp/s,'' is compared with the reference pattern Yp p/B- corresponding to this sample pattern, and the distance MDs of both patterns is compared. On the other hand, the rotation frequency f
A comparison is made based on the RMS value of the vibration amplitude at . Then, a trend display of the distance and RMS value is displayed to indicate whether the vibration mode is abnormal or only the amplitude is abnormal, and if there is an abnormality in the f2 distance or the RMS value S, a notification is issued.

[Effects of the Invention] According to the present invention (2), the probability that a slight change in the rotation speed of a rotating device such as a pump or an electric motor in a normal state is erroneously judged as abnormal is reduced. Reliability is improved.In addition, when the amplitude of vibration changes due to a change in the number of rotations of rotating equipment, it is determined whether the change is normal or not, and even in the case of an abnormal change, the vibration frequency can be determined. Since it can be determined whether the pattern shape (vibration mode) has changed, it is possible to provide useful information for plant operator judgment.

[Brief explanation of the drawing]

FIG. 1 is a graph showing the relationship between power spectral density Y and frequency f for detailed explanation of the present invention, and FIGS. 2 and 3 are analytical diagrams of FIG. The graph normalized by , Figure 4 shows the frequency e
A graph showing the relationship between f and the RMS value S of the amplitude, FIG. 5 is a block diagram showing one embodiment of the present invention, and FIG.
Figure Flowchart-C; h no. 1. . 1...Rotating equipment 2...Rotational speed meter 3...Signal human power device 4...Data storage device 5...Fourier analysis device 6...Material pattern correction device 7...Standard data storage Device 8... Comparison device 9.
...Output display device IO...Alarm generating device 11...
・Control device agent: Patent attorney Yoshiaki Inomata (and one other person). Figure 4 Figure 5 709 Figure 6

Claims (1)

[Claims] (1) A detection signal from a rotating device is sequentially analyzed to determine the power spectrum consistency t or coherence obtained for each frequency A sample value based on the γ-order vector f2, which is formed by the spectral density or coherence, is obtained, while the power spectral density based on the detection signal I of the rotating equipment in the normal state is determined by the 4th-order vector C2, which is formed by the coherence. In order to compare the previous S'L sample value and the corresponding reference value, the frequency value is normalized by the rotation frequency of the signal and the sample value due to the dispersion of 1 d is stored. Compare the sample pattern based on the sample value with the reference pattern based on the reference value by normalizing the reference value, and compare the vibration amplitude based on the effective value l2 of the signal amplitude with a threshold value. A method for monitoring rotating equipment characterized by displaying the results and issuing a warning when there is an abnormality in the rotating equipment. (2) Calculate the distance between the reference value and the sample value, and compare this distance with a pre-given threshold. 82 Reference pattern 1: Compare the corresponding sample pattern to determine the abnormality of the rotating equipment. Claim No. 1 in which the presence or absence is indicated
Method for monitoring rotating equipment as described in section. (8) The comparison result is displayed as a trend of the value obtained by comparing the vibration amplitude with the threshold value based on the distance between the reference value and the sample value and the effective value of the signal amplitude. A monitoring method for rotating equipment that displays whether there is any abnormality.