WO2019044745A1

WO2019044745A1 - Method and device for monitoring condition of rolling bearing

Info

Publication number: WO2019044745A1
Application number: PCT/JP2018/031517
Authority: WO
Inventors: 英之筒井; 正嗣北井
Original assignee: Ntn株式会社
Priority date: 2017-08-31
Filing date: 2018-08-27
Publication date: 2019-03-07

Abstract

The present invention estimates the remaining service life of grease sealed in a rolling bearing. A rolling bearing state monitoring method according to the present invention includes a step for producing a classifier by machine learning, and a step for monitoring the condition of rolling bearings. The step for producing a classifier includes steps (S11-S13) for calculating a feature amount for first through third vibration data, and a step (S14) for producing a classifier by machine learning of the feature amounts of the first through third vibration data. The first vibration data is vibration data of a rolling bearing in which grease having a first oil separation rate is sealed. The second vibration data is vibration data of a rolling bearing in which grease having a second oil separation rate, which is lower than the first oil separation rate, is sealed. The third vibration data is vibration data of a rolling bearing in which normal grease is sealed. The monitoring step includes a step for using the classifier to calculate the remaining service life until the oil separation rate of the grease reaches the first oil separation rate.

Description

Method and apparatus for monitoring state of rolling bearing

The present invention relates to a method and a device for monitoring the state of a rolling bearing.

Conventionally, a method of monitoring the state of a rolling bearing is known. For example, Japanese Patent Application Laid-Open No. 2004-347401 (Patent Document 1) discloses a rolling bearing having a simple configuration efficiently and accurately by using a background noise component having no peak such as a rotational speed proportional component in a frequency spectrum. A diagnostic method of a rolling bearing capable of determining the presence or absence of an abnormality is disclosed.

Unexamined-Japanese-Patent No. 2004-347401

Japanese Patent Laid-Open No. 2004-347401 (Patent Document 1) discloses a configuration in which it is determined that the grease sealed in the rolling bearing has deteriorated when the frequency spectrum of the vibration data of the rolling bearing exceeds a threshold. .

In order to properly determine the grease replacement time, it is necessary to estimate the remaining life of the grease according to the degree of deterioration. However, in Japanese Patent Application Laid-Open No. 2004-347401 (Patent Document 1), the estimation of the remaining life of the grease according to the degree of deterioration is not taken into consideration.

The present invention has been made to solve the above-mentioned problems, and its object is to estimate the remaining life of the grease sealed in the rolling bearing.

The rolling bearing state monitoring method according to the present invention includes the steps of creating a classifier by machine learning and monitoring the rolling bearing state. The step of creating a classifier includes the steps of calculating feature quantities of the first to third vibration data, and machine learning of the feature quantities of the first to third vibration data to create a classifier. The first vibration data is vibration data of the rolling bearing in which the first oil separation rate grease is enclosed. The second vibration data is vibration data of a rolling bearing in which a grease having a second oil separation rate smaller than the first oil separation rate is sealed. The third vibration data is vibration data of a rolling bearing in which normal grease is sealed. The operation time of the rolling bearing until the oil separation rate of the normal grease becomes the first oil separation rate is the first operation time. The operation time of the rolling bearing until the oil separation rate of the normal grease becomes the second oil separation rate is a second operation time which is shorter than the first operation time. The step of monitoring the state includes the steps of calculating the feature quantity of the fourth vibration data, calculating the first to third fitness factors, and calculating the remaining life of the grease sealed in the rolling bearing. . The fourth vibration data is vibration data of the rolling bearing measured during monitoring. The first to third relevance factors are calculated by the classifier. The first to third adaptation rates are adaptation rates between the feature quantities of the fourth vibration data and the feature quantities of the first to third oscillation data. The remaining life of the grease sealed in the rolling bearing is calculated using the first and second operation times and the first to third fitness factors. The remaining life is the operating time of the rolling bearing until the oil separation rate of the grease sealed in the rolling bearing becomes the first oil separation rate.

According to the method of monitoring the state of a rolling bearing according to the present invention, it is possible to estimate the remaining life of the grease sealed in the rolling bearing by using the classifier created by machine learning of the feature amount of the vibration data.

It is a figure which shows collectively the sectional view of the rolling bearing monitored by the state monitoring apparatus which concerns on embodiment, and a state monitoring apparatus. It is a functional block diagram of the state monitoring apparatus of FIG. It is a flowchart which shows the outline | summary of the estimation process of the remaining life of the grease performed by a control part. It is a flowchart which shows concretely the flow of the process performed in a learning period. It is a figure which shows together (a) vibration data (acceleration data) of oil separation level 2 (oil separation rate 70%), and (b) STFT image. It is a figure which shows together the time chart of (a) vibration data (acceleration data) of oil separation level 1 (oil separation rate 30%), and (b) STFT image. It is a figure which shows the time chart of normal (a) vibration data (acceleration data), and the (b) STFT image collectively. It is a flowchart which shows concretely the flow of the process performed in a monitoring period. It is a time chart of the matching rate of the STFT image of vibration data measured during monitoring, and the STFT image of each oil separation level. It is a figure which shows the relationship between the driving | running time of a rolling bearing, and the remaining life estimated by the state monitoring apparatus which concerns on embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding portions are denoted by the same reference characters and description thereof will not be repeated.

FIG. 1 is a view showing a cross-sectional view of a state monitoring device 1 and a rolling bearing 10 monitored by the state monitoring device 1 according to the embodiment. As shown in FIG. 1, the rolling bearing 10 includes an inner ring 12, an outer ring 14, a cage 16, and a plurality of rolling elements 18. Rolling bearing 10 includes, for example, self-aligning roller bearings, tapered roller bearings, cylindrical roller bearings, and ball bearings. The rolling bearing 10 may be a single row or multiple rows.

The inner ring 12 is a rotating ring which is fitted and fixed to the main shaft 11 and rotates integrally with the main shaft 11 in the direction of the arrow D. The outer ring 14 is a stationary ring disposed on the outer peripheral side of the inner ring 12.

The holder 16 is provided with a plurality of pockets for holding a plurality of rolling elements 18 at equal intervals. The cage 16 is disposed between the outer peripheral surface of the inner ring 12 and the inner peripheral surface of the outer ring 14 while holding the plurality of rolling elements 18. When the rolling element 18 rotates along the inner circumferential surface (track surface) of the outer ring 14 along with the rotation of the inner ring 12, the cage 16 together with the plurality of rolling elements 18 and the outer circumferential surface of the inner ring 12 and the inner circumferential surface of the outer ring 14 Rotate between

In order to form an oil film around metal components (for example, the inner ring 12, the outer ring 14, the cage 16, and the rolling elements 18) inside the rolling bearing 10, grease is used to inhibit metal-to-metal contact. Grc is enclosed.

The state monitoring device 1 monitors the state of the rolling bearing by measuring the physical quantity of the rolling bearing 10 by the vibration sensor. Examples of physical quantities of the rolling bearing 10 measured by the vibration sensor include acceleration, velocity, displacement, sound, AE (Acoustic Emission), and power.

As the operating time of the rolling bearing 10 increases, the grease Grc deteriorates and the oil separation rate of the grease Grc increases. When the oil separation ratio of the grease Grc increases, it becomes difficult to form a sufficient oil film around the rolling elements 18 and the like, and metal-to-metal contact is likely to occur inside the rolling bearing 10. If the metals are welded together by the frictional heat due to the contact between the metals, the rotation of the rolling bearing 10 becomes difficult.

The life (lubrication life) of the rolling bearing 10 due to the deterioration of the grease Grc is shorter than the life of the rolling bearing 10 due to damage (rolling fatigue) formed on the raceway surface by the plurality of rolling elements 18. In order to use the rolling bearing 10 for as long as possible, it is necessary to estimate what the remaining lubrication life (remaining life) is and to replace the grease Grc at an appropriate time.

Therefore, in the embodiment, the matching rate between a certain feature amount and a plurality of machine-learned feature amounts by machine learning the feature amounts of the vibration data of a rolling bearing in which greases having different oil separation rates are enclosed. Create a classifier that calculates By using the classifier, the remaining life of the grease sealed in the rolling bearing can be estimated.

FIG. 2 is a functional block diagram of the state monitoring device 1 of FIG. As shown in FIG. 2, the state monitoring device 1 includes a vibration sensor 20, a control unit 40, and a storage unit 50. The vibration sensor 20 measures vibration data of the rolling bearing 10 and outputs it to the control unit 40.

In the learning period, the control unit 40 machine-learns feature amounts of vibration data of the rolling bearing 10 in which greases having different oil separation rates are sealed, and creates a classifier. The classifier calculates the matching rate between a certain feature amount and each feature amount that has been machine-learned. Classifiers can include, for example, support vector machines, neural networks, naive Bayes, and decision trees. In the embodiment, a neural network is used as a classifier. The control unit 40 calculates the feature quantity of the vibration data measured at the sampling time in the monitoring period, and calculates the remaining life of the grease Grc using the classifier created in the learning period. The control unit 40 includes a computer such as a central processing unit (CPU).

The storage unit 50 stores, for example, vibration data, feature quantities calculated from the vibration data, and a classifier created by machine learning.

FIG. 3 is a flowchart showing an outline of the process of estimating the remaining life of grease performed by the control unit 40. Hereinafter, the step is simply described as S. As shown in FIG. 3, the control unit 40 creates a classifier at S1. S1 is a process performed in the learning period. The control unit 40 calculates the remaining life of the grease using a classifier at S2. S2 is a process performed in the monitoring period. The classifier created by machine learning may be created at the start of the monitoring period, and the learning period and the monitoring period may not be continuous.

FIG. 4 is a flowchart specifically showing the flow of processing (S1 in FIG. 3) performed in the learning period. The behavior of the rolling bearing 10 is not stable for a while after the rolling bearing 10 starts operation. Therefore, in the learning period, a break-in operation period (for example, 5 days) has elapsed since the operation of the rolling bearing 10 was started, and measurement is performed for a fixed period (for example, 1 month) with the behavior of the rolling bearing 10 stabilized. It is desirable that vibration data be used.

As shown in FIG. 4, the control unit 40 generates a short time Fourier transform (STFT) image of the vibration data of the rolling bearing in which the grease having the oil separation rate of oil separation level 2 is enclosed in S11, The feature amount corresponding to the oil separation level 2 is calculated, and the process proceeds to S12. The oil separation rate of oil separation level 2 is set as the oil separation rate of the grease which has almost exhausted the lubrication life and needs to be replaced. The oil separation rate at oil separation level 2 is, for example, 50% or more and 80% or less, and is 70% in the embodiment.

In S12, the control unit 40 calculates the STFT image of the vibration data of the rolling bearing in which the grease having an oil separation rate of oil separation level 1 is enclosed as a measurement corresponding to oil separation level 1, and the process proceeds to S13. Advance. The oil separation rate of oil separation level 1 is set as the oil separation rate of the grease, which is deteriorated due to the grease but the need for replacement is low because the lubrication life is maintained to some extent. The ratio of the oil separation rate of oil separation level 1 to the oil separation rate of oil separation level 2 is, for example, a value of 40% or more and 60% or less. In the embodiment, the oil separation rate at oil separation level 1 is 30%.

In S13, the control unit 40 calculates the STFT image of the vibration data of the rolling bearing in which the oil separation level 0 grease is enclosed as a feature corresponding to the oil separation level 0, and advances the process to S14. Oil separation level 0 grease is normal grease. A normal grease is a grease that hardly undergoes deterioration, for example, after the operation of a rolling bearing filled with unused grease, grease with an oil separation rate below the standard value, or unused grease is started. Grease within the threshold time. In the embodiment, the grease of oil separation level 0 is used as the grease within 5 hours after the start of operation of the rolling bearing in which the unused grease is sealed.

In S14, the control unit 40 machine-learns the STFT image of each grease of oil separation levels 0 to 2 to create a classifier, and advances the process to S15. The classifier calculates the matching rate of a certain STFT image with that of each oil separation level.

At S15, control unit 40 associates the STFT images of oil separation levels 0 to 2 with operation times L0 to L2 and stores them in storage unit 50, and ends the processing. The grease having the oil separation level 0 is a normal grease, so in the embodiment, the operating time L0 is set to 0. The operation time L1 is the operation time of the rolling bearing until the oil separation rate of the normal grease reaches the oil separation rate of oil separation level 1. The operating time L2 is the operating time of the rolling bearing until the oil separation rate of the normal grease reaches the oil separation rate of oil separation level 2. The operating time L1 may be the operating time when the rolling bearing is actually operated until the oil separation rate of unused grease reaches the oil separation rate of oil separation level 1, or the relationship between the operating time and the oil separation rate It may be the operating time obtained from The same applies to the operating time L2. In the embodiment, the operating time L1 is 1000 hours, and the operating time L2 is 2000 hours.

S11 to S13 may be performed before S14, and need not be performed in the order shown in FIG. Also, S11 to S13 may be performed not simultaneously but concurrently.

FIG. 5 is a diagram showing together (a) vibration data (acceleration data) of oil separation level 2 (oil separation rate 70%) and (b) STFT image. FIG. 6 is a diagram showing a time chart of (a) vibration data (acceleration data) of oil separation level 1 (oil separation rate 30%) and (b) STFT image together. FIG. 7 is a diagram showing together a normal (a) time chart of vibration data (acceleration data) and a (b) STFT image. The vibration data shown in FIGS. 5 to 7 is data obtained by measuring the vibration acceleration of an NTA angular contact ball bearing 7216 operating at a rotational speed of 1500 min ⁻¹ at a sampling speed of 50 kHz. In the measurement of the vibration data shown in FIGS. 5 to 7, the radial load and the axial load are both 1.3 kN, and the temperature is 150.degree.

In S11 of FIG. 4, FIG. 5A to FIG. 5B are calculated. In S12 of FIG. 4, FIG. 6A to FIG. 6B are calculated. In S13 of FIG. 4, FIG. 7A to FIG. 7B are calculated. In S14 of FIG. 4, the machine learning of FIG. 5 (b), FIG. 6 (b) and FIG. 7 (b) is performed.

The greases of

oil separation levels

1 and 2 may be greases obtained by actually operating the rolling bearings to be monitored, or greases having an artificially adjusted oil separation ratio OS. As a method of artificially adjusting the oil separation rate OS of grease, for example, a method of removing oil separated by a centrifugal separator from grease using the following formula (1), or heating grease by a thermostat or the like A method of evaporating the oil or separating the oil. In Formula (1), T0 is a thickener amount (%) of unused grease, and T1 is a thickener amount (%) of grease to be adjusted.

OS = (1-T0 / T1) × 100 (1)
The amount of thickener is measured by diluting the weighed grease with petroleum benzine, separating it with a centrifugal separator, removing the supernatant oil and petroleum benzine several times, The weight of the thickener may be measured to determine the proportion of the weight of thickener to the weight of grease.

In S11 and S12 of FIG. 4, the vibration data of the rolling bearing in which the grease of the enclosed amount G1 is enclosed assuming the weight change due to the oil separation at the set oil separation rate is referred to. The enclosed amount G1 can be determined, for example, from the following equation (2). In Formula (2), G0 is the enclosed amount (initial enclosed amount) of the grease enclosed at the time of the driving | operation start of a rolling bearing. The enclosed amounts G0 and G1 in the equation (2) are ratios of the volume of the enclosed grease to the total volume inside the rolling bearing in which the grease is enclosed. In the embodiment, the initial enclosed amount G0 is 27%.

G1 = G0 × (100−OS) / 100 (2)
FIG. 8 is a flowchart specifically showing the flow of the process (S2 in FIG. 3) performed in the monitoring period. The process shown in FIG. 8 is executed at each sampling time by a main routine (not shown).

As shown in FIG. 8, the control unit 40 calculates an STFT image of the vibration data (measurement data) measured at the sampling time in S21, and advances the process to S22. In step S22, the control unit 40 uses the classifier created in the learning period to calculate the relevance factors P0 to P2 between the STFT image of measurement data and each of three machine-learned STFT images, and advances the process to step S23. . The matching rate P0 is a matching rate between the STFT image of the measurement data and the STFT image of oil separation level 0. The matching rate P1 is a matching rate between the STFT image of the measurement data and the STFT image of oil separation level 1. The matching rate P2 is a matching rate between the STFT image of the measurement data and the STFT image of oil separation level 2. The sum of the relevance factors P0 to P2 is one.

In S23, the control unit 40 calculates the remaining life ΔL until the oil separation rate of the grease becomes the oil separation rate of oil separation level 2 using the following equation (3), and returns the process to the main routine.

ΔL = L2− (L0 × P0 + L1 × P1 + L2 × P2) (3)
If remaining life ΔL is smaller than the threshold, control unit 40 notifies the user of, for example, a message prompting replacement of grease.

Below, the result of the experiment which estimated the remaining life of the grease of a rolling bearing is shown using the state monitoring apparatus which concerns on embodiment. In the experiment, an angular contact ball bearing 7216 manufactured by NTN was operated at a rotational speed of 1500 min ^-1 , and vibrational acceleration of the angular contact ball bearing 7216 was measured at a sampling speed of 50 kHz for 20 seconds every two hours. Both radial load and axial load were 1.3 kN, and the temperature was 150 ° C. by heater heating. The oil separation rate of the grease enclosed inside the angular contact ball bearing 7216 was 70% when the operating time reached 2000 hours.

FIG. 9 is a time chart of the matching rate of the STFT image of vibration data measured during monitoring and the STFT image of each oil separation level. In FIG. 9, a curve C0 shows a time change of the matching rate P0 of the STFT image of vibration data measured during monitoring and the STFT image of oil separation level 0. The curve C1 shows the time change of the matching rate P1 of the STFT image of the vibration data measured during monitoring and the STFT image of oil separation level 1. A curve C2 shows the time change of the matching rate P2 of the STFT image of vibration data measured during monitoring and the STFT image of oil separation level 2.

As shown in FIG. 9, the operating time L10 (0 <L10 <500) is closest to the operating time L0 (0) associated with the oil separation level 0, and is then associated with the oil separation level 1 It is close to operating time L1 (1000). The actual state of the grease at the operation time L10 is closest to the state of the oil separation level 0 grease and next to the state of the oil separation level 1 grease. The values of the fitness factors P0 to P2 at the operating time L10 are 0.9, 0.1, and 0, respectively. From the magnitude relationship P0> P1> P2 of the fitness ratio P0 to P2, it is assumed that the state of the grease at the operating time L10 is closest to the state of the oil separation level 0 grease and next to the state of the oil separation level 1 grease It can be estimated. The actual condition of the grease in the operating time L10 is reflected in the adaptation rates P0 to P2 of the operating time L10.

The operating time L20 (1000 <L20 <1500) is closest to the operating time L1, and next to the operating time L2 (2000) associated with the oil separation level 2. The actual state of the grease at the operation time L20 is closest to the state of the oil separation level 1 grease and next to the state of the oil separation level 2 grease. The values of the fitness factors P0 to P2 at the operating time L20 are 0, 0.9, and 0.1, respectively. From the magnitude relationship P1> P2> P0 of the fitness ratio P0 to P2, the state of the grease at the operating time L20 is closest to the state of the oil separation level 1 grease and next to the state of the oil separation level 2 grease It can be estimated. The actual condition of the grease at the operating time L20 is reflected in the adaptation rates P0 to P2 of the operating time L20.

The operating time L30 (1500 <L30 <2000) is closest to the operating time L2, and is next closest to the operating time L1. The actual state of the grease at the operation time L30 is closest to the state of the oil separation level 2 grease and next to the state of the oil separation level 1 grease. The values of the fitness factors P0 to P2 at the operating time L30 are 0, 0.1, and 0.9, respectively. According to the relationship P2> P1> P0 of the fitness ratio P0 to P2, the state of the grease at the operating time L30 is closest to the state of the oil separation level 2 grease and next to the state of the oil separation level 1 grease It can be estimated. The actual condition of the grease in the operating time L30 is reflected in the adaptation rates P0 to P2 of the operating time L30.

FIG. 10 is a view showing the relationship between the operation time of the rolling bearing and the remaining life estimated by the condition monitoring device according to the embodiment. As shown in FIG. 10, the remaining life corresponding to the state of the grease which is deteriorated with the increase of the operation time is estimated.

In the embodiment, the case where the inner ring is a rotating ring and the outer ring is a stationary ring has been described. In the rolling bearing to be monitored by the state monitoring device according to the embodiment, the inner ring may be a stationary ring and the outer ring may be a rotating ring.

In the embodiment, the oil separation rate is divided into three levels, and the case where the STFT image of each level is machine-learned in the learning period has been described. The oil separation rate may be divided into four or more levels. Also, the feature amount of the vibration data may be other than the STFT image, and may be, for example, an effective value, a maximum value, a crest factor, a kurtosis, and a skewness.

As described above, according to the state monitoring device according to the embodiment, the remaining life of the grease sealed in the rolling bearing can be estimated by using the classifier created by machine learning of the feature amount of the vibration data.

It should be understood that the embodiments disclosed herein are illustrative and non-restrictive in every respect. The scope of the present invention is shown not by the above description but by the scope of claims, and is intended to include all modifications within the scope and meaning equivalent to the scope of claims.

1 condition monitoring device, 10 rolling bearing, 11 spindle, 12 inner ring, 14 outer ring, 16 cage, 18 rolling element, 20 vibration sensor, 40 control unit, 50 storage unit, Grc grease.

Claims

Creating a classifier by machine learning;
Monitoring the state of the rolling bearing,
The step of creating the classifier comprises
Calculating a feature amount of the first vibration data of the rolling bearing in which the first oil separation rate grease is enclosed;
Calculating a feature amount of the second vibration data of the rolling bearing in which the grease having the second oil separation rate smaller than the first oil separation rate is enclosed;
Calculating a feature amount of third vibration data of the rolling bearing in which normal grease is sealed;
Machine learning the feature quantities of the first to third vibration data to create the classifier;
The operation time of the rolling bearing until the oil separation rate of the normal grease becomes the first oil separation rate is a first operation time,
The operation time of the rolling bearing until the oil separation rate of the normal grease becomes the second oil separation rate is a second operation time shorter than the first operation time,
The step of monitoring the condition may
Calculating a feature of the fourth vibration data of the rolling bearing measured during monitoring;
Calculating, using the classifier, first to third fitness factors of the feature quantities of the fourth vibration data and the feature quantities of the first to third vibration data;
The remaining life until the oil separation rate of the grease enclosed in the rolling bearing reaches the first oil separation rate is calculated using the first and second operation times and the first to third compatibility rates. And monitoring the state of the rolling bearing.
The sum of the first to third precisions is 1,
In the step of calculating the remaining life, the remaining life is obtained by subtracting the sum of the values obtained by multiplying the first and second adaptation rates by the first and second operating times, respectively, from the first operating time. The state monitoring method of the rolling bearing according to claim 1, which is calculated as
The method of monitoring the condition of the rolling bearing according to claim 1 or 2, wherein the grease of the first oil separation rate and the grease of the second oil separation rate are artificially adjusted greases.
The grease of the first oil separation rate is a grease obtained by operating the rolling bearing in which the normal grease is enclosed for the first operation time,
The rolling bearing according to claim 1 or 2, wherein the grease having the second oil separation rate is a grease obtained by operating the rolling bearing in which the normal grease is enclosed for the second operation time. State monitoring method.
The state monitoring of the rolling bearing according to any one of claims 1 to 4, wherein the feature amounts of the first to fourth vibration data respectively include data obtained by subjecting the first to fourth vibration data to short time Fourier transform. Method.
The state of the rolling bearing according to any one of claims 1 to 5, wherein the first to third vibration data are vibration data measured after lapse of a break-in operation period after operation of the rolling bearing is started. How to monitor.
A state monitoring device that monitors the state of the rolling bearing using the method of monitoring the state of the rolling bearing according to any one of claims 1 to 6.