JPS5842949A - Monitoring method for body in periodic motion - Google Patents

Abstract

PURPOSE:To monitor a body in periodic motion without reference to equipment size, a load, etc., by comparing the dispersive normalized value of the even- numbered moment of the probability density function of oscillation data with the normal logical value of the body in periodic motion, and thus detecting a fault. CONSTITUTION:The oscillation of reduction gears 1 for a body in periodic motion is detected by an oscillation detector 4 and sampled at specified intervals by a sampling circuit 5, whose output is A/D-converted and then stored in a storage device 6. On the basis of the storage contents, a computing device 7 performs arithmetic according to equalitiesIand II to calculate the dispersive normalized value K of the even-numbered moment of the probability density function of the oscillation data which corresponds to the extent of a normal distribution. Then, a comparing circuit 9 compares said calculated value with a logical dispersive normalized value K as a constant, such as three, of the body in periodic motion stored in a memory 8 and when their difference is great, a decision on the occurrence of a fault is made to operate an alarm device 10, etc. This constitution monitors the body in periodic motion without reference to the equipment size, load, etc. Further, P(x):x is the probability density function.

Description

[Detailed Description of the Invention] The present invention relates to a method for monitoring objects that perform periodic motion such as rotation, such as bearings and gears. The present invention relates to a method of detecting this using a universal criterion that is unrelated to Fi, such as the size and load status of equipment equipped with periodic moving bodies.

In general, if abnormalities such as scratches on parts, eccentricity of the rotating shaft, or poor lubrication occur in periodic moving bodies such as bearings and gears, and equipment equipped with these, if the abnormalities occur, such as parts of the bearings and gears, etc. This causes not only damage but also failure and destruction of the entire equipment equipped with them. Therefore, accurately grasping the extent of such abnormality is an extremely important issue in the maintenance and management of equipment including bearings, gears, etc. Conventionally, in order to understand the extent of such abnormalities, a senna is placed in a periodic moving body, and the RMS (Root Mean 5quare) value or power spectrum integral value is determined from the data obtained from the output signal, and the equipment is determined by analyzing them. The plan was to diagnose and monitor the situation. The RMS value is 1+, similar to the value defined by the following CI) formula, xI:
Time series data (izl, 2''·N) N: The data power spectrum integral value SP is a value defined by the following equation (2).

5(f): Power spectrum value at frequency f [f,, f1]: Integral interval These values vary depending on the size of the equipment, the magnitude of the load, and the rotation speed, etc. - This is a universal judgment. Unable to set a standard, individual normal data was released #! A#-i This was done in consideration of certain circumstances, and when an abnormality occurred in periodic moving bodies such as gears and bearings, the size of the equipment,
It is an object of the present invention to provide a method for monitoring a periodic moving body, such as a load, which detects an abnormality using a universal criterion that is not affected by the conditions of the e-monitoring body.

The method for monitoring a periodic moving body according to zero shot #14 is to calculate the higher-order moment of a probability density function formed by a group of sampling values obtained by subjecting the vibration of a periodic moving body to constant period '''cr sampling as the sampling value. The value normalized by the variance of the group is compared with the theoretical value of the normalized higher-order moment related to the sampled value group when the periodic body is normal, and based on the comparison result, it is determined whether the periodic body is abnormal. It is characterized by detecting °.

First, the principle of the present invention will be described. If the monitored object is normal, the probability density function (P
robustness density function
(P, D, F (hereinafter omitted)) are generally considered to have a normal distribution, but the moment of PLD, F, - can be mentioned as a means of quantifying this disturbance, which is the main part of . The latter is selected because the odd-order moments indicate the symmetry of the distribution, and the even-order moments indicate the spread of the distribution. In addition, since the dispersion is large for days and evening groups with large amplitudes, the disturbance can be quantified using an even-order moment with an order larger than the second order, and a value normalized by the dispersion, for example, the index KK shown in the following formula. become

Similar, X: Time series data Kugen: X's P, D, F.

E: average operating voice,: fourth moment l! :Dispersion of X ” When the monitored object is normal, its P, D, and F have a theoretical normal distribution, and the value of K is theoretically 3.0.

Therefore, regardless of the size of the equipment, load status, etc., the value of K is used as the criterion, and the degree of separation from the value of K of 8.0 is IK-8,
01, the degree of abnormality of the monitored object is grasped. That is, when the object to be monitored is abnormal, -3.01 becomes large in Ul, so the degree of abnormality in the object to be monitored can be grasped from this value.

Next, the method of the present invention will be explained based on drawings showing examples thereof. FIG. 1 is a schematic diagram showing the implementation state of the method of the present invention, in which a gear 1a having a predetermined gear ratio of a reducer IKI/i,
The shaft 1g, which is fitted into the center hole of the gear 1m, is supported by bearings lc and ld placed opposite to the housing of the reducer l, and the shaft 1g, which is fitted into the center hole of the gear 1b, is built-in. 1h is supported by bearings le and 1fK, which are provided oppositely to the housing of the reducer 1.

The shaft 1g is connected to the output shaft of the morph 2, and its rotation is decelerated by the gears 1a and 1bK and transmitted to the shaft 1hK, and further transmitted to the load pump 3 via the pulley and Peyre. There is. bearing lc.

A vibration detection device 114 that detects the vibration and converts it into an electric signal is attached to the outer ring of ld, le, and if. The analog data is sampled and converted into digital data, and stored in the storage device 6.
The data will be stored sequentially. The digital data stored in the storage device 6 is input to the calculation device 7, and the value of K is determined for each detection unit based on the following equation, which is a modification of equation (3) to handle digital values.

Ni-Leaving・-! I' (x(t)-x) dt#'
TO=-! -*' j (x(t6+ (i 1)Δ)
-M)'(4),4,'f1. ．． 1 Similar, t: Ordinal number representing the sampling time point The average value of is also determined based on the following equations (5) and (6), which are modified to handle digital values.

(5) In addition, the storage device it8 stores data related to the normal level and the life and death level, and the data and the value of K obtained by the calculation device 7 are input to the comparison circuit 9, and the comparison circuit 9 As a result of comparison in circuit 9, if the monitored object is determined to be abnormal, v! An electrical signal is sent to the alarm device 10, and *m is emitted.

The effectiveness of the method of the present invention was confirmed by creating artificial scratches (pitting scratches) on the gears of a speed reducer using the apparatus described above, and the results are shown in FIGS. 2 and 3. That is, FIG. 2 shows the value of #'iK, and FIG. 3 shows the probability density function. In the figure, A is normal data, BJ/'i gear has small scratches,
Status data C indicates data for a gear with large scratches.

During normal operation, the K value #ia shows a value close to O, but when the gear has thick scratches, it can exceed Fi8.0 and show an abnormally high value (approximately 8.6). It took a minute.

Next, we will explain the results of confirming the effectiveness of the method of the present invention by creating defects in the bearings and gears in the reducer shown in FIG. 4, and causing misalignment in the coupling. '
@4 The reduction gear shown in Figure 4 consists of a gear aS (number of teeth = 15) supported by bearings B, , B, and attached to nine shafts S, and a shaft supported by bearings B, , B. Gear Gl (number of teeth = 90) and gear G3 (number of teeth:
15) and bearing B1. B.

111] S, supported by gear G,
Using (m = 87), (gear G1 and gear G,
Gears □G and G are meshed with each other to transmit deceleration to the front, and are used in the reduction gear.

A vibration detection device was installed in the outer ring of the bearing, and the results of confirming the effectiveness of the method of the present invention were shown in the first part.

From this result, it was found that the value of K was abnormally high except in the case where the bearing outer ring had a spot defect, and it was found that the abnormality could be clearly identified by the method of the present invention.

The value of K for determining normality and abnormality cannot be specified unconditionally as it depends on the accuracy of use of the equipment, but as a general guideline, the following predetermined values obtained empirically are considered normal. It is possible to distinguish between abnormalities and abnormalities.

2.5 A to A 8.5: Normal 8, 5 < K <, 4.0: Abnormal (alarm required), K
≦4.0: Abnormality (4iL disassembly examination) In addition, it goes without saying that the reliability will be higher than K by using the conventional RMS value analysis, and furthermore, in the above example, the 4th order of the probability density function Although the value of K was determined based on the moment), it goes without saying that a higher-order moment may be used.

As described in detail above, in the case of the present invention, a probability density function constituted by a group of sampling values obtained by sampling the vibration of a periodic body at a constant period is based on a value normalized by a fraction of the group of sampling values. Since it distinguishes between normal and abnormal periodic moving objects, it is possible to monitor periodic moving objects that are not affected by the size of the equipment equipped with periodic moving objects, the magnitude of the load, and the rotational speed, etc. There is no need to accumulate data, and universal abnormality judgment criteria can be established, making it highly useful.

[Brief explanation of the drawing]

Figure 1 is a schematic diagram showing the implementation status of the method of the present invention, Figures 2 and 3 are graphs showing the results of confirming the effects of the method of the present invention, and Figure 4 is a diagram showing the results of confirming the effects of the method of the present invention. FIG. 2 is a schematic diagram showing a reduction gear used. 1-...Reducer 1a, 1b...Gear 1c, ld
, 1eeIf...Bearing 2...Motor 3
...Pump 4...Vibration detection device f Gt, G
t, Ga, c, ・[Table Bt * J e Bs
e B4 * 13s e Bg...Bearing special
Applicant: Sumitomo Metal Industries Co., Ltd. Agent Patent Attorney Noboru Kono

Claims (1)

[Claims] 1. A value obtained by normalizing the higher-order moment of a probability density function formed by a group of sampling values obtained by sampling the vibration of one body at a constant period by the variance of the group of sampling values, and the periodic movement body. is compared with the theoretical value of the normalized higher-order moment related to the sampled value group in the case where the sampling value group is positive and negative, and an abnormality in the periodic moving body is detected based on the comparison result.7. .