Roller bearing diagnostic method, roller bearing diagnostic program, recording medium, and information processing device

Abstract

To provide a roller bearing diagnosis method which can easily detect abnormalities in roller bearings that may lead to abnormalities in rollers and determine the degree of the abnormalities.SOLUTION: A roller bearing diagnosis method for a belt conveyor according to one embodiment of the present disclosure includes a process of measuring radiation sound from a belt conveyor in operation, a process of calculating at least one value of an indicator representing energy fluctuation and an indicator representing sound sharpness about the radiation sound measured in the measuring process, and a step of determining abnormalities in the bearing based on the value calculated in the calculating process.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a roller bearing diagnosis method, a roller bearing diagnosis program, a recording medium, and an information processing apparatus.

Background Art

[0002] A belt conveyor is widely used as a system for conveying an object to be conveyed. Generally, a belt conveyor includes a belt for conveying the placed object to be conveyed, a roller for driving the belt, and a roller for supporting the belt. These rollers may develop abnormalities due to aging fatigue, entry of foreign substances such as dust, etc., resulting in a decrease in the efficiency of conveying the object to be conveyed. Methods for detecting such roller abnormalities are known (Japanese Patent Application Laid-Open No. 2016-060556).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In Patent Document 1, the acceleration of the belt is measured by an acceleration sensor arranged on the side surface of the belt, and a non-rotating roller and its position are obtained from the measurement data. With this method, although it is possible to detect a roller having a terminal abnormality of being unable to rotate, there is a possibility that a roller having an abnormality before reaching the terminal cannot be detected.

[0005] In view of such circumstances, an object of the present disclosure is to provide a roller bearing diagnosis method that can easily detect an abnormality of a roller bearing that may cause an abnormality of a roller and can determine the degree of the abnormality.

Means for Solving the Problems

[0006] The roller bearing diagnosis method of the belt conveyor according to one aspect of the present disclosure for solving the above problems includes a step of measuring the radiated sound from the belt conveyor during operation, a step of calculating at least one value of an index representing the energy fluctuation and an index representing the sharpness of the sound for the radiated sound measured in the measuring step, and a step of determining the abnormality of the bearing based on the value calculated in the calculating step.

Effect of the Invention

[0007] The roller bearing diagnosis method of the belt conveyor of the present disclosure can easily detect the abnormality of the roller bearing that can cause the abnormality of the roller and can determine the degree of the abnormality.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Mode for Carrying Out the Invention

[0009] [Description of the Embodiment of the Present Disclosure] First, the embodiments of the present disclosure will be listed and described.

[0010] (1) The roller bearing diagnosis method of a belt conveyor according to one aspect of the present disclosure includes a step of measuring radiated sound from an operating belt conveyor, a step of calculating at least one value of an index representing energy fluctuation and an index representing sharpness of sound for the radiated sound measured in the measuring step, and a step of determining bearing abnormality based on the value calculated in the calculating step.

[0011] The roller bearing diagnosis method of the belt conveyor (hereinafter also referred to as the diagnosis method) calculates at least one value of an index representing energy fluctuation and an index representing sharpness of sound from the measured radiated sound of the belt conveyor, and determines bearing abnormality based on this calculated value. Therefore, the influence of noise generated from the belt conveyor or the like can be suppressed, the abnormality of the roller bearing, which is the main source of abnormality of the roller of the belt conveyor, can be easily detected, and the degree of abnormality can be easily determined from the width, magnitude, etc. of the calculated value.

[0012] (2) In the above (1), the index representing the energy fluctuation may be at least one of impact, fluctuation intensity, and roughness. By the index representing the energy fluctuation being at least one of impact, fluctuation intensity, and roughness, the index can be easily calculated from the measured radiated sound.

[0013] (3) In the above (1) or (2), the index representing the sharpness of the sound may be sharpness. By the index representing the sharpness of the sound being sharpness, the index can be easily calculated from the measured radiated sound.

[0014] (4) In any one of the above (1) to (3), in the measuring step, the microphone may be moved parallel to the running direction of the belt of the belt conveyor to measure the radiated sound. By doing so, the radiated sound can be easily measured.

[0015] (5) The roller bearing diagnosis program for a belt conveyor according to one aspect of the present disclosure causes an information processing device to execute an element that calculates at least one value of an index representing energy fluctuation and an index representing sharpness of sound for data obtained by measuring radiated sound from an operating belt conveyor, and an element that determines abnormality of a bearing based on the calculated element.

[0016] According to the roller bearing diagnosis program for the belt conveyor (hereinafter also referred to as the program), at least one value of an index representing energy fluctuation and an index representing sharpness of sound is calculated from the measured radiated sound of the belt conveyor, and based on this calculated value, abnormality of the bearing can be determined. Therefore, it is possible to suppress the influence of noise generated from a belt conveyor or the like and easily detect abnormality of a roller bearing, which is a main source of abnormality of a roller of the belt conveyor, and it is possible to easily determine the degree of abnormality from the width, magnitude, etc. of the calculated value.

[0017] (6) A recording medium according to one aspect of the present disclosure has the program recorded thereon.

[0018] Since the program is recorded on the recording medium, abnormality of a roller bearing can be easily detected by incorporating it into or externally connecting it to an information processing device that executes the program.

[0019] (7) An information processing device according to one aspect of the present disclosure has the recording medium incorporated therein or externally connected thereto.

[0020] Since the recording medium is incorporated in or externally connected to the information processing device, abnormality of a roller bearing can be easily detected by executing the program.

[0021] [Details of the Mode for Carrying Out the Present Disclosure] Hereinafter, the present disclosure will be described in detail with reference to the drawings as appropriate. Note that the drawings are for illustrative purposes, and each configuration (each member) is conceptually depicted, and the shape, scale, etc. may be different from the actual ones.

[0022] <Belt conveyor roller bearing diagnosis method> The diagnosis method includes a step of measuring the radiated sound from a belt conveyor during operation, a step of calculating at least one value of an index representing the energy fluctuation and an index representing the sharpness of the sound for the radiated sound measured in the measuring step, and a step of determining the possibility of bearing abnormality based on the value calculated in the calculating step.

[0023] The diagnosis method can be used, for example, for diagnosing the roller bearings of the belt conveyor 1 as shown in FIG. 1. The belt conveyor 1 of the present embodiment has an endless belt 10 on which an object S to be conveyed is placed, a driving roller 21 that drives the belt 10, and a plurality of support rollers 22 that support the belt 10. The driving roller 21 rotates by a driving device (not shown) to run the belt 10 in one direction. The plurality of support rollers 22 support the belt 10 while rotating by the driven belt 10. That is, the support roller 22 is a driven roller. The diagnosis method can be used for any of the driving roller 21 and the support roller 22 (hereinafter, these rollers are also collectively referred to as "roller 20").

[0024] The material of the belt 10 is not particularly limited, and it may be made of metal, resin, rubber, or a combination formed by laminating these materials. The total length (circumference length) and width of the belt 10 are not particularly limited, and may be appropriately set according to the size of the object S to be conveyed, the conveying distance, and the like.

[0025] As shown in FIGS. 2 and 3, the roller 20 includes a substantially cylindrical roll portion 210, a pair of substantially disk-shaped housing portions 220 that seal the openings at the ends of the roll portion 210, a bearing 230 disposed at substantially the center of each of the pair of housing portions 220, and a shaft 240 supported by the bearing 230.

[0026] The materials of the roll part 210 and the housing part 220 are not particularly limited as long as they can support the belt 10, and may be metal, resin, or those formed by laminating or combining these. The diameters (outer diameters) of the roll part 210 and the housing part 220 are not particularly limited and may be set as appropriate.

[0027] As shown in FIG. 4, the bearing 230 has an outer ring 231, an inner ring 232, and a plurality of balls 233 disposed between the outer ring 231 and the inner ring 232. The bearing 230 of the present embodiment is a so-called radial ball bearing. The outer ring 231 is fitted to the housing part 220, and the shaft 240 is fitted to the inner ring 232. By rolling a plurality of balls 233 disposed between the outer ring 231 and the inner ring 232, the roll part 210 and the housing part 220 rotate with respect to the shaft 240. The material of each member of the bearing 230 is preferably metal, but may be resin.

[0028] The shaft 240 is a substantially rod-shaped member that supports the roll part 210 and the housing part 220 via the bearing 230. The material of the shaft 240 is preferably metal, but may be resin.

[0029] 〔Measuring step〕 In the measuring step, the radiated sound from the belt conveyor 1 during operation is measured. That is, the radiated sound of the belt conveyor 1 in a state where the roller 20 is rotating and the belt is being driven is measured. This measurement may be performed in a state where the belt 10 is conveying the object to be conveyed S, or in a state where the belt 10 is not conveying the object to be conveyed S.

[0030] It is preferable to measure the radiated sound by moving the microphone parallel to the traveling direction of the belt 10 of the belt conveyor 1. Specifically, it is preferable to provide a moving body that moves substantially parallel to the belt 10 and attach the microphone to this moving body to measure the radiated sound. The operator may carry the microphone and measure the radiated sound while moving substantially parallel to the belt 10. That is, it is preferable that the microphone is not attached to the belt conveyor or fixed at a predetermined position separated from the belt conveyor. By doing so, the radiated sounds of a plurality of rollers 20 included in the belt conveyor 1 can be easily and efficiently measured with one microphone. The movement of the microphone does not need to be exactly parallel to the traveling direction of the belt 10, and it is sufficient if it is along the belt 10.

[0031] The microphone is not particularly limited as long as it can pick up the radiated sound, but it is preferable to use a directional microphone. By using a directional microphone, the accuracy of identifying the roller 20 having an abnormality can be improved. In addition, it is possible to suppress the collection of sounds (noise) other than the radiated sound.

[0032] 〔Calculation step〕 In the calculation step, at least one value of an index representing the energy fluctuation and an index representing the sharpness of the sound is calculated for the radiated sound measured in the measurement step. That is, this diagnostic method diagnoses the abnormality of the roller using these sound quality evaluation indexes. Such a diagnosis has the following advantages compared to the conventional diagnosis.

[0033] As a conventional diagnostic method, for example, as shown in FIG. 5, there is a diagnostic method that focuses on the difference (magnitude) in sound energy. When the sound energy levels obtained by analyzing the measured sound of the roller are averaged for each degree of abnormality (damage degree) of the roller, a positive correlation is observed between the above levels and the above damage degree, but the variation (spread of the distribution) of the levels at the same damage degree is large. Therefore, it may be difficult to obtain the sound level of an arbitrarily selected roller and estimate the damage degree from this level. For example, at a level of 95 dB where damage is suspected, as shown in FIG. 5, it may correspond to any of the damage degrees 1 indicating minor damage to the damage degree 3 indicating severe damage. One of the reasons why it becomes difficult to discriminate the damage degree in this way is that the level may be dominated by the sound caused by the rotating system. That is, even if the roller is normal, the operating sound is generated when the conveyor operates, and the sound level caused by the damage of the roller is relatively small compared to the level of the operating sound. Therefore, it is necessary to extract the component caused by the damage of the roller instead of the operation of the conveyor.

[0034] The damage leading to the loss of function of the roller is mainly caused by the bearing. As a conventional diagnostic method for detecting bearing damage, a method that focuses on the specific frequency (natural frequency) emitted by the damaged bearing is known. This is obtained geometrically by calculating the period in which the rolling element (ball) passes through the damage when there is a scratch on the inner ring of the bearing, etc., from the specifications of the rolling bearing and the operating conditions (rotation speed), and is widely used in the diagnosis of the rotor bearings of rotating machinery.

[0035] In the diagnostic method of the present disclosure, instead of the above-mentioned natural frequency, diagnosis is performed by capturing changes that appear near the natural vibration frequency of the roller housing. This is due to the following reasons. The first reason is that the rotational speed may not be constant. In the case of a rotating machine, if the rotational speed is constant, it rotates at that rotational speed. However, a conveyor roller is a driven roller (a system in which the belt moves at a constant speed and the roller is rotated by the friction between the belt and the roller). In order to support a single belt with multiple rollers, even if there is some slippage in some of the rollers (in extreme cases, it may not rotate), the function as a conveyor is not lost. For this reason, when looking at a single roller, the rotational speed may not necessarily be constant. The second reason is that there are multiple mechanisms for roller damage (loss of function). The loss of roller function is often due to bearing damage. As examples of the loss of roller function, there are a pattern in which the bearing is first damaged by aging deterioration such as wear and then progresses (a case assumed in general bearing diagnosis), and a pattern in which the bearing is rotated by the belt in a state where it is difficult to rotate due to the influence of foreign substances (for example, dust), and the damage progresses. Depending on these patterns, the way the exciting force that determines the radiated sound is generated is different. On the other hand, in any pattern, ultimately the sound is radiated from the roller housing, and by capturing changes near the natural vibration frequency of this housing, the degree of roller damage can be detected.

[0036] Among conventional diagnostic methods, there is also a method that focuses on changes that appear in a specific frequency band, and a technique of extracting and analyzing the frequency band corresponding to the natural vibration frequency of the housing can also be considered. However, the natural vibration frequency of the housing may change depending on the structure of the roller, the load from the belt, etc., and may also change due to the aging deterioration of the roller. Furthermore, as shown in FIG. 6, the frequency characteristics of damaged rollers may differ depending on individual rollers. Therefore, it is not easy to specify in advance the natural vibration frequency of the housing of a damaged roller and set the frequency band corresponding to that natural vibration frequency.

[0037] Therefore, in the present diagnostic method, at least one of the sound quality evaluation indices representing the energy variation and the sound quality evaluation index representing the sharpness (pitch sharpness) of the sound is used so that changes appearing near the natural frequency can be captured without previously identifying the natural frequency of the housing. The sound quality evaluation index representing the energy variation can preferably capture an abnormal roller that particularly emits a fluctuating sound. The sound quality evaluation index representing the energy variation is preferably at least one of Impulsiveness, Fluction Strength, and Roughness.

[0038] The sound quality evaluation index representing the sharpness of the sound can preferably capture an abnormality in which the high-frequency component increases. The sound quality evaluation index representing the sharpness of the sound is preferably Sharpness.

[0039] Since the above sound quality evaluation indices may or may not be suitable for detecting abnormalities depending on the diagnostic target and the type of abnormality, it is preferable to perform diagnosis by combining a plurality of the above sound quality evaluation indices. Each of the sound quality evaluation indices will be described in detail below, but these sound quality evaluation indices may be calculated using a commercial program.

[0040] (Impulsiveness) Impulsiveness is an index corresponding to the subjective impulsiveness of the sound, and can quantitatively evaluate the energy variation in the radiated sound. First, I’ is obtained by the following formula (1).

[0041] [Equation] Here, k j is a weighting coefficient. e J is an excitation function and corresponds to the waveform to be diagnosed. y(e J ) is a compression function, which is approximately linear for small amplitudes and a function similar to a power function of 0.15 for large amplitudes.

[0042] I’ compares the one obtained by giving an envelope waveform to an impact vibration waveform and normalizing it with the absolute value of the value. If it is an impact steady signal, the numerator becomes 0. Note that the weighting coefficient k j is a constant. When used in the diagnosis method, it does not have to use the same value for normal and abnormal rollers. Substitute I’ obtained by Equation 1 into Equation 2 below to calculate the impact I.

[0043] I = 0.055556×I’+heavi(I’-1.8)× (1.271367×I’-2.288461)+ heavi(I’-7)×(2.288461-0.326923×I’) ····(2) Here, heavi(x) is a function such that heavi(x)=0 (x<0), 1 (x≧0).

[0044] The impact can be equivalent to the one calculated by subtracting the average value from the measured waveform of the above-mentioned radiation sound, extracting only the variation, and dividing this by the above-mentioned average value, so as not to depend on the amplitude of the above-mentioned waveform, but rather to calculate the ratio of the variation to the total energy of the above-mentioned waveform.

[0045] (Variation intensity) The variation intensity is an index corresponding to the strength of the subjective sense of sound variation. The variation intensity graphs the measured variation waveform pattern of the sound pressure of the above-mentioned radiation sound as shown in Figure 5. Envelope processing is performed on the variation of the sound pressure level of this graph, and the variation level and variation period are calculated to obtain the drop ΔL and the modulation frequency f mod and. Calculate this ΔL and f mod for each critical band frequency and substitute them into Equation 3 below to calculate the variation intensity F.

[0046]

Equation

[0047] (Roughness) Roughness is an index corresponding to the subjective roughness of sound. Multiple calculation formulas are known for roughness. The calculation formula used for calculating roughness is not particularly limited, but in this embodiment, ΔL and fmod are obtained in the same manner as the above-described fluctuation intensity, and are substituted into the following formula 4 for calculation. Roughness is likely to show a correlation with the abnormality of the roller 20 when the fluctuation period is short (the fluctuation frequency is high).

[0048]

Number

[0049] (Sharpness) Sharpness is an index corresponding to the subjective sharpness or shrillness of sound. As the damage of the bearing 230 progresses, the damage and the discontinuity of the sliding surface increase, thereby exciting the resonance of the structure. Usually, since the natural frequencies of the bearing 230 and the housing part 220 are higher than the rotational component of the roll part 210, the sound in the high frequency band tends to be dominant. Particularly in the case of the driven roller (support roller 22), the housing part 220 is often used until the end, and the natural frequency of the housing part 220 is likely to change due to the progress of damage. For example, when diagnosing based on the difference in frequency between abnormal and normal conditions, although the difference in frequency increases due to damage in the initial stage, the difference in the above frequency may be suppressed due to the decrease in the natural frequency. In sharpness, the abnormality of the roller 20 can be detected by focusing on the center of gravity of the waveform rather than the difference in specific frequencies.

[0050] There are multiple calculation models for sharpness, but in this embodiment, it is calculated using the following formula 5.

[0051]

Number

[0052] z is in the unit of [Bark] and represents the band of the critical band. The integration range corresponds to the audible range expressed in terms of the critical band width. For example, g'(z) is 1 when the critical band is from 0 Bark to 16 Bark or less (below the midrange), and increases as the critical band increases above 16 Bark. Sharpness approximately corresponds to the centroid of the spectrum, and the higher the frequency of the centroid, the sharper (more acute) the sound.

[0053] Impact, fluctuation intensity, roughness, and sharpness may be diagnosed individually, but the diagnostic accuracy can be improved by combining two or more of them. Abnormalities in the bearing may cause differences in the generation of the excitation force of the radiated sound depending on the content thereof, that is, the degree of wear, flaw, deformation, etc., and the process (formation mechanism) in which they are formed. By combining two or more of impact, fluctuation intensity, roughness, and sharpness, highly accurate diagnosis can be performed regardless of the content of the abnormality.

[0054] [Determination step] In the determination step, based on the values calculated in the above calculation step, the abnormality of the bearing is determined. The determination is made not only on the presence or absence of an abnormality in the roller 20 but also on the degree of the abnormality.

[0055] In this diagnostic method, since the sound quality evaluation index is used, the degree of abnormality can be determined from the values calculated by the above formulas 1 to 5. As a method for determining the degree of abnormality, for example, it may be determined by the magnitude of the value of the calculated sound quality evaluation index, or the degree of abnormality in the roller having an abnormality and the value of the sound quality evaluation index of the abnormal roller are previously tabulated, and a control value (threshold value) is set from this data, and the presence or absence and / or degree of abnormality are determined based on the comparison result between the value of the sound quality evaluation index calculated from the measured radiated sound and the control value.

[0056] In addition, the above management value may be set based on at least one of TPR (True Positive Rate) and FPR (False Positive Rate). TPR is [the number of abnormal products determined to be abnormal] / [the total number of actual abnormal products], and FPR is [the number of normal products determined to be abnormal] / [the total number of actual normal products]. It is most preferable to determine a management value such that TPR is 1 (100%) and FPR is 0 (0%), but there may be cases where such a management value cannot be obtained. Therefore, for example, in a diagnosis where detection omission of abnormal products is not allowed, a diagnosis for the purpose of a certain degree of screening, etc., appropriate target values for at least one of TPR and FPR are set according to the purpose of the diagnosis, and the management value may be set so as to obtain the target value. As an example, in the case of a diagnosis where detection omission of abnormal products is not allowed, setting the management value so that TPR is 1 (100%) can be mentioned. Also, as another example, in the case of a diagnosis for the purpose of a certain degree of screening, setting the management value so that TPR is 0.8 (80%) can be mentioned. In this way, by setting the management value based on at least one of TPR and FPR, it becomes possible to perform a diagnosis according to the purpose.

[0057] In addition, as the determination of the degree of abnormality, for example, the degree of abnormality is divided into four levels, level 1 is a normal product, level 2 is a moderately abnormal product, level 3 is an abnormal product close to the end stage, and level 4 is an abnormal product in the end stage (abnormal product that cannot be used). The degree of abnormality of each roller 20 may be allocated from the values calculated by the above formula 1 to the above formula 5.

[0058] Since the diagnosis method calculates a sound quality evaluation index from the measured radiated sound of the belt conveyor and determines the abnormality of the bearing based on this calculated value, it can suppress the influence of noise generated from the belt conveyor, etc., and can easily detect the abnormality of the roller bearing, which is the main source of abnormality of the rollers of the belt conveyor, and can also easily determine the degree of abnormality.

[0059] <Belt Conveyor Roller Bearing Diagnosis Program> The program is one that programs the above-described diagnosis method. By being programmed so as to be processable by an information processing apparatus such as a personal computer and recorded on a recording medium, it is possible to easily diagnose abnormalities in the roller bearing regardless of the skill level of the operator.

[0060] <Recording medium> The program is recorded on a recording medium. Since the program is recorded on the recording medium, by being built in or externally connected to an information processing apparatus that executes the program, it is possible to easily diagnose abnormalities in the roller bearing.

[0061] <Information processing apparatus> The recording medium is built in or externally connected to an information processing apparatus. Since the recording medium is built in or externally connected to the information processing apparatus, by executing the program, it is possible to easily diagnose abnormalities in the roller bearing regardless of the skill level of the operator.

[0062] [Other embodiments] The above embodiments do not limit the configuration of the present invention. Therefore, based on the description in this specification and common technical knowledge, it is possible to omit, replace, or add components of each part of the above embodiments, and all of them should be interpreted as belonging to the scope of the present invention.

[0063] The configuration of the roller is not limited to that described above. For example, the roll part and the pair of housing parts may be integrally formed.

[0064] The bearing is not limited to a radial ball bearing, and may be a roller bearing or a thrust bearing. Further, it may be a bearing having a seal part so that the balls are not exposed.

[0065] The diagnosis method may be used in combination with diagnosis based on other data, such as sound energy, sound pressure, frequency characteristics, etc.

[0066] The acoustic quality evaluation index representing the variation in sound energy is not limited to those described above. For example, kurtosis, which is an index used for evaluating the impact of sound and represents the sharpness of the statistical distribution, DLF (Depth of Loudness Fluctuation) related to sound variation, etc. may be used.

[0067] Also, for the purpose of removing noise to clarify the characteristic quantities of the diagnosis target, frequency filters such as high-pass, low-pass, and band-pass may be applied.

Example

[0068] Hereinafter, the present disclosure will be further described by way of examples, but the present disclosure is not limited to these examples.

[0069] In each of the following examples, the sound near the roller during conveyor operation was measured during the operator's patrol of the belt conveyor, and the sound was analyzed to evaluate the usefulness of each index. Specifically, the state of the roller with a known damage degree in advance was represented by damage degrees 1 to 4 (1; normal, 2; mid-stage damage, 3; late-stage damage, 4; end-stage damage), and the measurement and analysis results were associated with the damage degree. The evaluation of the measurement and analysis results was based on about 100 cases with damage degrees 1 to 3, excluding rollers without damage (not corresponding to damage degrees 1 to 4) and end-stage damage (damage degree 4, which is significantly different in nature from others even without considering differences in analysis methods). ArtemiS SUITE 12.1 of Head Acoustics was used for calculating various acoustic quality evaluation indexes.

[0070] TPR and FPR were used as evaluation indicators to show the usefulness of diagnosis. The reason is that when looking at a distribution diagram like the level in Fig. 5 (conventional diagnostic method), if there is a strong correlation between the damage degree (horizontal axis) and the feature quantity (vertical axis), it can be judged as a significant feature quantity. However, in actual data, there are multiple plots that overlap, so it is difficult to quantitatively determine which one is superior. Here, TPR (True Positive Rate) is defined as the percentage of [the number of actually defective products discriminated as defective] / [the total number of actually defective products], and FPR (False Positive Rate) is defined as the percentage of [the number of actually normal products discriminated as defective] / [the total number of actually normal products]. Here, since the roller (damage degree 4) with obvious abnormalities that is essential for replacement during operation has been excluded from the data in advance, for the data of abnormal rollers with minor to severe damage degrees (damage degrees 1 to 3) that remain, when the control value (threshold) is set so that the TPR is about 80%, it was evaluated from the perspective of how low the FPR would become. In this example, those with a damage degree of 1 were counted as normal products, and those with damage degrees of 2 and 3 were counted as defective products.

[0071] [Example 1] In Example 1, as an inventive example, an index representing the variation in the energy of the radiated sound, specifically, the impact and the variation intensity were calculated to diagnose the abnormality of the bearing. Also, as a comparative example, the abnormality diagnosis based on the sound pressure level of the above radiated sound and the 1 / 3 octave band analysis was also performed. In the 1 / 3 octave band analysis, representative frequency bands of 2000 Hz and 4000 Hz were set. The diagnosis was made by comparing each diagnosis with the value of FPR when setting a control value such that the TPR becomes 0.8 (80%) using TPR and FPR. The results are shown in Table 1.

[0072] [Table 1]

[0073] In terms of the sound pressure level, the FPR is 43%, and nearly half of the normal products are diagnosed as defective products. In the diagnosis by the 1 / 3 octave band analysis, the accuracy is improved compared to the diagnosis of the sound pressure level, but the accuracy varies depending on the frequency band to be analyzed.

[0074] In contrast, in the diagnosis at the variation intensity in the inventive example, without specifying the frequency band, it is possible to ensure the same diagnostic accuracy as the 1 / 3 octave band analysis result in the 4000 Hz band. In the diagnosis of impact, the accuracy is further improved, and normal products (damage degree 1) were not determined as abnormal products (damage degree 2 or damage degree 3).

[0075] [Example 2] In Example 2, as an inventive example, an index representing the sharpness of sound, specifically, the sharpness of the radiated sound was calculated for bearing abnormality diagnosis. Also, as a comparative example, abnormality diagnosis was performed by 1 / 3 octave band analysis in the frequency band of 8000 Hz in the higher frequency range. The diagnosis was compared for each diagnosis with the value of FPR when setting the control value so that TPR becomes 0.8 (80%) using TPR and FPR. The results are shown in Table 2.

[0076]

Table 2

[0077] The diagnostic accuracy of sharpness was higher than the 1 / 3 octave band analysis at 4000 Hz but lower than the 1 / 3 octave band analysis at 8000 Hz. This is considered to be due to the fact that it is an index that calculates loudness for each critical band and adds them together, as can be seen from the formula of sharpness. However, for bearing abnormalities, the frequency band where differences in frequency characteristics appear varies depending on the content of the damage. Since sharpness does not require presetting the frequency band and has good affinity with the operator's (measurer's) sense of hearing, it is considered that easy and efficient abnormality diagnosis can be performed.

Industrial Applicability

[0078] The method for diagnosing the roller bearing of the belt conveyor of the present disclosure can easily detect abnormalities in the roller bearing, and thus can be effectively used for the maintenance of the belt conveyor.

Explanation of Signs

[0079] 1 Belt conveyor 10 Belt 20 Roller 21 Driving roller 22 Supporting roller 210 Roll part 220 Housing 230 Bearing 231 Outer ring 232 Inner ring 233 Ball 240 Shaft S Object to be conveyed

Claims

1. A step of measuring the radiation noise from an operating belt conveyor; A step of calculating at least one value of an index representing the energy fluctuation and an index representing the sharpness of the sound for the radiation noise measured in the above measuring step; A step of determining the abnormality of the bearing based on the value calculated in the above calculating step A method for diagnosing a roller bearing of a belt conveyor, comprising the above steps.

2. The roller bearing diagnosis method according to claim 1, wherein the index representing the energy fluctuation is at least one of impact, fluctuation intensity, and roughness.

3. The roller bearing diagnosis method according to claim 1, wherein the index representing the sharpness of the sound is sharpness.

4. The roller bearing diagnosis method according to any one of claims 1 to 3, wherein in the above measuring step, the microphone is moved parallel to the traveling direction of the belt of the belt conveyor to measure the above radiation noise.

5. An element for calculating at least one value of an index representing the energy fluctuation and an index representing the sharpness of the sound for the data obtained by measuring the radiation noise from an operating belt conveyor; An element for determining the abnormality of the bearing based on the above calculated element A roller bearing diagnosis program for a belt conveyor that causes an information processing device to execute the above steps.

6. A recording medium on which the roller bearing diagnosis program according to claim 5 is recorded.

7. An information processing device in which the recording medium according to claim 6 is built-in or externally connected.

Images