Vibration analysis system and vibration analysis method

Abstract

A vibration analysis system (100) includes a signal input unit (202) that receives an input of a vibration signal detected by a sensor mounted on a moving object, a strength calculation unit (204) that calculates a plurality of signal strengths corresponding to a plurality of frequency bands by analyzing the vibration signal corresponding to the object, a first distance calculation unit (206) that calculates a first Mahalanobis distance of a first signal space constituted by the plurality of signal strengths with respect to a first unit space, a center-of-gravity calculation unit (208) that calculates two-dimensional center-of-gravity data indicating a center-of-gravity position of the plurality of signal strengths, a second distance calculation unit (210) that calculates a second Mahalanobis distance of a second signal space constituted by the center-of-gravity data with respect to a second unit space, and an abnormality prediction unit (212) that predicts a period of abnormality occurrence of the object based on the first Mahalanobis distance and the second Mahalanobis distance.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to a vibration analysis system and a vibration analysis method. BACKGROUND

[0002] In the past, as a method for checking abnormalities of a machine, a method of determining whether or not a device has an abnormality by detecting a signal due to an abnormal vibration in the operation of the device has been known.

[0003] For example, in Japanese Patent Application Publication No. 2019-35585 (Patent Literature 1), an abnormality determination process based on a prescribed proximity method is applied to each data of a new analysis result data group corresponding to an existing representative data group representing an existing analysis result data group, and the abnormality determination process calculates an index value indicating a degree of abnormality based on a distance between data to determine whether or not it is abnormal.

[0004] PRIOR ART DOCUMENTS

[0005] PATENT LITERATURE

[0006] Patent Literature 1: Japanese Patent Application Publication No. 2019-35585 SUMMARY

[0007] PROBLEMS TO BE SOLVED BY THE INVENTION

[0008] In Patent Literature 1, selection of representative data for abnormality detection from a large amount of data as a state monitoring result of an object device is discussed. However, neither disclosure nor suggestion is made regarding any structure for predicting a period of abnormality for predictive maintenance of a device or the like.

[0009] An object of an aspect of the present disclosure is to provide a vibration analysis system and a vibration analysis method capable of predicting a period of occurrence of an abnormality of an object by analyzing a vibration state of the object.

[0010] SOLUTION TO PROBLEM

[0011] The vibration analysis system according to one embodiment includes a signal input unit configured to accept input of a vibration signal detected by a sensor mounted to a subject in motion; a strength calculation unit configured to calculate a plurality of signal strengths corresponding to a plurality of frequency bands by analyzing the vibration signal corresponding to the subject; a first distance calculation unit configured to calculate a first Mahalanobis distance of a first signal space constituted by the plurality of signal strengths with respect to a first unit space set in advance; a center-of-gravity calculation unit configured to calculate two-dimensional center-of-gravity data indicating a center-of-gravity position of the plurality of signal strengths calculated by the strength calculation unit; a second distance calculation unit configured to calculate a second Mahalanobis distance of a second signal space constituted by the two-dimensional center-of-gravity data with respect to a second unit space set in advance; and an abnormality prediction unit configured to predict an abnormality occurrence timing at which the subject is abnormal based on the first Mahalanobis distance and the second Mahalanobis distance.

[0012] Preferably, the first unit space is constituted by a plurality of signal strengths corresponding to a plurality of frequency bands calculated by analyzing a vibration signal corresponding to the subject at a normal time. The second unit space is constituted by two-dimensional center-of-gravity data indicating a center-of-gravity position of the plurality of signal strengths constituting the first unit space.

[0013] Preferably, the abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than a first threshold value is predicted to be a future closer than an abnormality occurrence timing in a case where the first Mahalanobis distance is not calculated to be equal to or higher than the first threshold value.

[0014] Preferably, the abnormality prediction unit predicts that the subject is abnormal after a number of days from when the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value.

[0015] Preferably, the abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value and the second Mahalanobis distance is calculated to be equal to or higher than a second threshold value is predicted to be a future closer than an abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value and the second Mahalanobis distance is not calculated to be equal to or higher than the second threshold value.

[0016] Preferably, the abnormality prediction unit predicts that the subject is abnormal after a number of hours from when the second Mahalanobis distance is calculated to be equal to or higher than the second threshold value.

[0017] Preferably, the vibration analysis system further includes an output control unit configured to output first warning information in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value, and output second warning information having a warning level higher than a warning level of the first warning information in a case where the second Mahalanobis distance is calculated to be equal to or higher than the second threshold value.

[0018] Preferably, the output control unit causes a display to display time-series data of the first Mahalanobis distance and time-series data of the second Mahalanobis distance.

[0019] The vibration analysis method according to other embodiments includes the steps of accepting input of a vibration signal detected by a sensor mounted on a subject in motion, calculating a plurality of signal intensities corresponding to a plurality of frequency bands, respectively, by analyzing the vibration signal corresponding to the subject, calculating a first Mahalanobis distance of a first signal space constituted by the plurality of signal intensities with respect to a first unit space set in advance, calculating two-dimensional center-of-gravity data indicating a center-of-gravity position of the plurality of signal intensities calculated, calculating a second Mahalanobis distance of a second signal space constituted by the two-dimensional center-of-gravity data with respect to a second unit space set in advance, and predicting an abnormality occurrence period in which the subject is abnormal based on the first Mahalanobis distance and the second Mahalanobis distance.

[0020] Effects of Invention

[0021] According to the present disclosure, it is possible to predict an abnormality occurrence period of a subject by analyzing a vibration state of the subject. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a diagram for explaining an outline of the system.

[0023] Figure 2 is a block diagram showing an example of the overall structure of the vibration analysis system.

[0024] Figure 3 is a block diagram showing an example of the hardware structure of the analysis device.

[0025] Figure 4 is a flowchart showing an example of the preparation process.

[0026] Figure 5 is a diagram showing an example of a data set of signal intensities of each frequency band.

[0027] Figure 6 is a diagram showing an example of a data set of center-of-gravity positions.

[0028] Figure 7 is a flowchart showing an example of the analysis process.

[0029] Figure 8 is a flowchart showing an example of the first abnormality prediction process.

[0030] Figure 9 is a diagram showing time series data of Mahalanobis distances based on multi-dimensional data.

[0031] Figure 10 is a flowchart showing an example of the second abnormality prediction process.

[0032] Figure 11 is a diagram showing time series data of Mahalanobis distances based on two-dimensional data.

[0033] Figure 12 is a flowchart showing an example of the trend analysis process.

[0034] Figure 13 is a diagram showing an example of the layout of a user interface screen.

[0035] Figure 14 is a functional block diagram of the analysis device. DETAILED DESCRIPTION

[0036] Hereinafter, an embodiment of the present application will be described with reference to the drawings. In the following description, the same reference numerals are assigned to the same parts. Their names and functions are also the same. Thus, detailed description thereof will not be repeated.

[0037] <SYSTEM CONFIGURATION>

[0038] Figure 1 is a diagram for explaining an outline of the system 1000. Referring to Figure 1 , the system 1000 is a system for predicting an abnormality of an object (hereinafter, also simply referred to as "object") in which a vibration occurs in operation, by analyzing a signal of the vibration. Hereinafter, the object will be described as a pump, but is not limited thereto, and the system 1000 can be applied to any object in which a vibration (or sound) occurs in operation. For example, the system 1000 can be applied to an abnormality prediction of a portion that vibrates due to a vibration from a motor, a vibrating body, or the like.

[0039] The system 1000 includes a vibration analysis system 100, a plurality of sensors 30, a terminal device 40, a network 50, and a plurality of pumps 70. The vibration analysis system 100 performs vibration analysis of the pumps 70. The vibration analysis system 100 includes an analysis device 10 and a sensor unit 20. The sensor unit 20 is electrically connected to the plurality of sensors 30. In the system 1000, two sensor units 20 are connected to the analysis device 10, but can be a structure in which three or more sensor units 20 are connected to the analysis device 10, or one sensor unit 20 is connected to the analysis device 10. Each sensor unit 20 can be electrically connected to one sensor 30. Each sensor unit 20 can be electrically connected to a plurality of sensors 30 respectively installed in the plurality of pumps 70.

[0040] The sensor 30 is installed in the pump 70, and acquires a detection signal (vibration signal) detected due to a vibration or sound of the pump. The analysis device 10 performs vibration analysis of the pump 70 based on the vibration signal input from the sensor 30 via the sensor unit 20. The analysis device 10 is configured to be able to communicate with the terminal device 40 via the network 50. The analysis device 10 transmits a vibration analysis result or the like to the terminal device 40.

[0041] Typically, the analysis device 10 has a configuration following a general-purpose computer architecture, and various processes described later are realized by executing a pre-installed program by a processor. The analysis device 10 is, for example, a laptop PC (Personal Computer). However, the analysis device 10 can be another device (for example, a desktop PC, a tablet terminal device) as long as it can execute the functions and processes described below.

[0042] The network 50 includes various networks such as the Internet. The network 50 can employ a wired communication method, or another wireless communication method such as a wireless LAN (local area network).

[0043] The terminal device 40 is, for example, a portable tablet terminal device. However, the terminal device 40 can be realized by a smartphone, a desktop PC (Personal Computer), or the like, without being limited thereto. Furthermore, the vibration analysis system 100 according to the present embodiment is configured by a separate type device in which the analysis device 10 and the sensor unit 20 are separate, but can be configured by a unitary type device in which the analysis device 10 and the sensor unit 20 are unitary.

[0044] Figure 2 is a block diagram showing an example of the overall configuration of the vibration analysis system 100. Referring to Figure 2 , the vibration analysis system 100 includes the analysis device 10 and the sensor unit 20.

[0045] The sensor 30 connected to the sensor unit 20 is a sensor capable of detecting a signal of vibration, sound, and the like, and is, for example, configured by an acceleration sensor using an organic piezoelectric element. Furthermore, the sensor 30 can be configured by another type (for example, a servo type) of acceleration sensor, or various other sensors, as long as it is a sensor capable of detecting a signal of vibration, sound, and the like.

[0046] In a case where the signal obtained by the sensor 30 is a charge signal, a charge converter is provided between the sensor 30 and the vibration analysis system 100. In this case, the charge converter converts the charge signal from the sensor 30 into a voltage signal and outputs it to the vibration analysis system 100. Furthermore, in a case where the sensor 30 has a function of converting a charge signal into a voltage signal, the charge converter is not needed.

[0047] The sensor unit 20 converts the vibration signal acquired from the sensor 30 (or the charge converter) into a signal that can be processed by the analysis device 10. Specifically, the sensor unit 20 includes a filter 21, an amplifier 22, and an A / D converter 23.

[0048] The filter 21 is an analog filter that removes a noise component from the vibration signal output from the sensor 30. The filter 21 is constituted by a low-pass filter, a high-pass filter, or the like.

[0049] The amplifier 22 amplifies the analog signal output from the filter 21 to a prescribed multiple, and outputs the amplified signal to the A / D converter 23.

[0050] The A / D converter 23 converts the signal input from the amplifier 22 from an analog signal to a digital signal at a prescribed sampling frequency. The A / D converter 23 outputs the signal obtained by the digital conversion to the analysis device 10.

[0051] Figure 3 is a block diagram showing an example of the hardware structure of the analysis device 10. Referring to Figure 3 , the analysis device 10 includes a processor 101, a memory 103, a display 105, an input device 107, a signal input interface (I / F) 109, and a communication interface (I / F) 111. These components are connected in a manner capable of performing data communication with each other.

[0052] Typically, the processor 101 is an arithmetic processing unit such as a CPU (Central Processing Unit), an MPU (Multi Processing Unit), or the like. The processor 101 controls the operation of each component of the analysis device 10 by reading out and executing a program stored in the memory 103. More specifically, the processor 101 realizes each function of the analysis device 10 by executing the program.

[0053] The memory 103 is realized by a RAM (Random Access Memory), a ROM (Read-Only Memory), a flash memory, a hard disk, or the like. The memory 103 stores a program or the like executed by the processor 101.

[0054] The display 105 is, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, or the like. The display 105 can be integrated with the analysis device 10, or can be constituted independently of the analysis device 10.

[0055] The input device 107 accepts an operation input to the analysis device 10. The input device 107 is realized by, for example, a keyboard, a button, a mouse, or the like. Alternatively, the input device 107 can be realized as a touch panel.

[0056] The signal input interface 109 relays data transmission between the processor 101 and the sensor unit 20. The signal input interface (I / F) 109 accepts input of a vibration signal from the sensor 30 via the sensor unit 20. Specifically, the signal input interface 109 accepts input of a digital signal from the A / D converter 23.

[0057] The communication interface 111 relays data transmission between the processor 101 and the terminal device 40 or the like. As a communication method, for example, a wireless communication method based on Bluetooth (registered trademark), wireless LAN (Local Area Network), or the like is used. Further, as a communication method, a wired communication method such as USB (Universal Serial Bus) can also be used.

[0058] <Abnormality prediction method>

[0059] An outline of the abnormality prediction method according to the present embodiment will be described. The abnormality prediction method includes a preparation process of preparing reference data, and an analysis process of analyzing an abnormal state of the pump 70.

[0060] (Preparation process)

[0061] In the preparation process according to the present embodiment, for example, a vibration state of the pump 70 at the initial stage of an operation start is measured. The pump 70 at the initial stage of the operation start is in a state of a new product, and thus a vibration state of the pump 70 at a normal time is measured as a reference. However, instead of the pump 70, a pump of the same kind as the pump 70 in a normal state can be prepared separately, and a vibration state of the pump can be measured as a reference.

[0062] Figure 4 is a flowchart showing an example of the preparation process. Typically, each of the following steps is realized by the processor 101 of the analysis device 10 executing a program saved in the memory 103.

[0063] Referring to Figure 4 , the processor 101 acquires a vibration signal output from the sensor 30 via the sensor unit 20 (step S10). Specifically, the processor 101 acquires a vibration signal corresponding to the pump 70 at a normal time (a vibration signal indicating a vibration state of the pump 70) from the sensor 30.

[0064] The processor 101 performs octave analysis on the vibration signals accumulated for a prescribed time (e.g., several tens of mseconds to several hundreds of mseconds) (step S12). In the present embodiment, 1 / 3 octave analysis is used. Thus, each vibration signal is separated into, for example, 48 frequency bands from 0.4 Hz to 20 kHz by a 1 / 3 bandpass filter, and the signal intensity (vibration intensity) is averaged for each frequency band (i.e., band). In the following description, the signal intensity obtained by averaging in the frequency band is also simply referred to as "signal intensity of the frequency band".

[0065] The processor 101 stores, for each frequency band, the signal intensity of the frequency band corresponding to the pump 70 at the normal time as reference data R to the memory 103 (step S14).

[0066] Figure 5 is a graph showing an example of a data set of the signal intensity of each frequency band. Referring to Figure 5 In the data set 310, the signal intensity L of each frequency band fl to fn (where n is a natural number, n < m) is included with respect to each time Tl to Tm (where m is a natural number). In the case where each vibration signal is separated into 48 frequency bands, n = 48. For example, the data set 310 includes the signal intensities Ll_l to Ll_n of each frequency band fl to fn at the time Tl, and the signal intensities Lm_l to Lm_n of each frequency band fl to fn at the time Tm. The reference data R includes, for example, the signal intensities L of each frequency band fl to fn at each time Tl to Tn. In this case, the period of the times Tl to Tn corresponds to the initial period of the operation.

[0067] Referring again to Figure 4 , the processor 101 sets a unit space Ul in the Mahalanobis-Taguchi method (MT method) using the reference data R (step S16). Specifically, the processor 101 sets the data of the signal intensities L of each frequency band fl to fn at each time Tl to Tn as the unit space Ul. The unit space Ul is used in the first abnormality prediction process described later.

[0068] The processor 101 calculates barycentric data of two dimensions indicating the barycentric position of the signal intensity of each frequency band (step S18). Specifically, the processor 101 generates a data set shown in Figure 6 based on the data set 310.

[0069] Figure 6 is a graph showing an example of a data set of the barycentric position. Referring to Figure 6 In the data set 320, the barycentric position Gx of the frequency band and the barycentric position Gy of the signal intensity are included with respect to the times Tl to Tm. The barycentric data (barycentric positions Gx, Gy) of two dimensions are generated for each time Tl to Tm. The barycentric position Gx is expressed by the following equation (1), and the barycentric position Gy is expressed by the following equation (2).

[0070] [Num 1]

[0071]

[0072] [Num 2]

[0073]

[0074] f i represents the i-th frequency band, L i represents the signal intensity of the i-th frequency band, S represents the sum of the signal intensities of all frequency bands. By the above (1) and (2), the data set 320 composed of two-dimensional data (center-of-gravity positions Gx, Gy) is generated on the basis of the data set 310 composed of n-dimensional (multi-dimensional) data (signal intensities L of the respective frequency bands fl to fn). In the data set 320, for example, the center-of-gravity positions Gx, Gy at the time Tl are represented by the center-of-gravity positions Gf_1, GL_1, respectively, and the center-of-gravity positions Gx, Gy at the time Tm are represented by the center-of-gravity positions Gf_m, GL_m, respectively.

[0075] Referring back to Figure 4 , the processor 101 sets the unit space U2 in the MT method using the center-of-gravity data in the action-start initial period (for example, the times Tl to Tn) (step S20). Specifically, the processor 101 sets the two-dimensional center-of-gravity data (center-of-gravity positions Gx, Gy) of the respective times Tl to Tn as the unit space U2. The unit space U2 is used in the second anomaly prediction process described later.

[0076] (Analysis Process)

[0077] The analysis process includes a first anomaly prediction process and a second anomaly prediction process that predict a period of abnormality occurrence of the vibration state of the pump 70, and a trend analysis process that analyzes a future trend of the vibration state of the pump 70.

[0078] Figure 7 is a flowchart showing an example of the analysis process. Referring to Figure 7 , the processor 101 acquires the vibration signal output from the sensor 30 via the sensor unit 20 (step S30). Specifically, the processor 101 acquires the vibration signal corresponding to the pump 70 in the normal period (for example, the period after the end of the action-start initial period) from the sensor 30.

[0079] The processor 101 performs octave analysis on the vibration signal accumulated for a prescribed time (step S32). The processor 101 stores the signal intensity of each frequency band corresponding to the pump 70 to the memory 103 for each frequency band (step S34). Specifically, the signal intensity of each frequency band in the pump 70 at a certain time Ts is stored in the form of a data set 310. Here, the series of signal intensities Ls_1 to Ls_n of the frequency bands f1 to fn in the pump 70 at the time Ts are also referred to as signal intensity data Ps. Further, the period from the time T1 to the time Tn corresponds to the initial period of the operation, and thus the period in general corresponds to the period from the time Tn+1 to the time Tm. Therefore, the time Ts is any time from the time Tn+1 to the time Tm.

[0080] The processor 101 uses the signal intensity data Ps and the reference data obtained in the preparation process to execute a first abnormality prediction process (step S40), a second abnormality prediction process (step S60), and a trend analysis process (step S70). These processes can be executed in parallel or sequentially.

[0081] [First Abnormality Prediction Process]

[0082] Figure 8 is a flowchart illustrating an example of the first abnormality prediction process. In the first abnormality prediction process, the data set 310 composed of multidimensional data is used. Referring to Figure 8 , the processor 101 calculates the Mahalanobis distance MD1 (hereinafter, also simply referred to as "distance MD1") of the signal space X1s composed of a plurality of signal intensities with respect to the unit space U1 set in step S16 of Figure 4 (step S41). The signal space X1s is composed of a plurality of signal intensities Ls_1 to Ls_n at the time Ts (i.e., the signal intensity data Ps).

[0083] The processor 101 determines whether the distance MD1 is equal to or greater than a threshold value Th1 (for example, 5) (step S43). In the case where the distance MD1 is less than the threshold value Th1 (NO in step S43), the processor 101 ends the first abnormality prediction process. In the case where the distance MD1 is equal to or greater than the threshold value Th1 (YES in step S43), the processor 101 predicts that the pump 70 will be abnormal in the near future, outputs an abnormality alarm (step S45), and ends the first abnormality prediction process. Typically, the abnormality alarm is displayed on the display 105. Further, the abnormality alarm can also be a structure that outputs a sound via a speaker.

[0084] Here, the reason why the processor 101 makes the prediction as described above will be described with reference to Figure 9 . Figure 9This is a graph showing the time series data of Mahalanobis distance based on multidimensional data. Graph 410 shows the time series data of Mahalanobis distance MDx1 in a reference pump of the same type as pump 70. Therefore, it can be said that the Mahalanobis distance MDx1 in graph 410 shows the same trend as the Mahalanobis distance MD1 in pump 70.

[0085] According to curve 410, the Mahalanobis distance MDx1 of the signal space relative to the unit space first exceeded the threshold Th1 (=5) at time Ta1. Subsequently, at times Ta2, Ta3, and Ta4, the Mahalanobis distance MDx1 exceeded the threshold Th1, and finally, at time Ta5, the reference pump malfunctioned. The period from time Ta1 to time Ta5 was 4.5 days, the period from time Ta2 to time Ta5 was 3 days, the period from time Ta3 to time Ta5 was 1.5 days, and the period from time Ta4 to time Ta5 was 8.5 hours. Therefore, it can be understood that the malfunction occurred several days after the Mahalanobis distance MDx1 first exceeded the threshold Th1.

[0086] Therefore, if the Mahalanobis distance MD1 is calculated to be above the threshold Th1, the processor 101 predicts that pump 70 will malfunction in the near future. Specifically, the processor 101 can also predict that pump 70 will malfunction several days after the initial calculation of the Mahalanobis distance MD1 above the threshold Th1. In this case, the processor 101 can also output a high-level alarm to notify that pump 70 may malfunction several days later.

[0087] [Second Anomaly Prediction Process]

[0088] Figure 10 This is a flowchart illustrating an example of the second anomaly prediction process. (Refer to...) Figure 10 The processor 101 calculates two-dimensional centroid data representing the centroid positions of the signal strength in each frequency band (step S61). Specifically, the processor 101 calculates the two-dimensional centroid data (i.e., centroid positions Gx, Gy) based on the signal strength data Ps calculated in step S32 using equations (1) and (2).

[0089] Processor 101 calculates the signal space X2s, which is composed of two-dimensional centroid data, relative to the data in the signal space X2s. Figure 4 The Mahalanobis distance MD2 (hereinafter also simply referred to as "distance MD2") of the unit space U2 is set in step S20 (step S63). The signal space X2 is composed of the centroid positions Gx and Gy (i.e., Gf_s and GL_s) at time Ts.

[0090] The processor 101 determines whether the distance MD2 is equal to or greater than a threshold value Th2 (for example, 5) (step S65). In a case where the distance MD2 is less than the threshold value Th2 (NO in step S65), the processor 101 ends the second abnormality prediction process. In a case where the distance MD2 is equal to or greater than the threshold value Th2 (YES in step S65), the processor 101 predicts that the pump 70 will be abnormal in the near future, outputs an abnormality alarm (step S67), and ends the second abnormality prediction process.

[0091] Here, the reason why the processor 101 predicts as described above will be described with reference to Figure 11 Figure 11 is a graph showing time-series data of the Mahalanobis distance based on the two-dimensional data. The graph 420 shows time-series data of the Mahalanobis distance MDx2 in a reference pump of the same kind as the pump 70. Thus, it can be said that the Mahalanobis distance MDx2 of the graph 420 shows the same tendency as the Mahalanobis distance MD2 in the pump 70.

[0092] According to the graph 420, the Mahalanobis distance MDx2 first exceeds the threshold value Th2 (= 5) at the time Tb1. Thereafter, the reference pump is abnormal at the time Tb2. The period from the time Tb1 to the time Tb2 is 5.5 hours. It can be said from this that the abnormality occurs several hours after the Mahalanobis distance MDx2 first exceeds the threshold value Th2.

[0093] Thus, in a case where the Mahalanobis distance MD2 equal to or greater than the threshold value Th2 is calculated, the processor 101 predicts that the pump 70 will be abnormal in the near future. In particular, the processor 101 can also predict that the pump 70 will be abnormal several hours after the Mahalanobis distance MD2 equal to or greater than the threshold value Th2 is first calculated. In this case, the processor 101 can also output an abnormality alarm of a high warning level for notifying that the abnormality is likely to occur several hours later.

[0094] [Tendency analysis process]

[0095] Figure 12 is a flowchart showing an example of the tendency analysis process. With reference to Figure 12 , the processor 101 calculates a difference H obtained by subtracting the representative data from the signal intensity data Ps (step S71).

[0096] The reference data R is constituted by a data group including the signal intensities L of the frequency bands fl to fn at each time Tl to Tn. For example, the processor 101 extracts the signal intensities Ln_1 to Ln_n of the frequency bands fl to fn at a certain time Tn in the data group as the representative data. Then, the processor 101 calculates the difference H obtained by subtracting the representative data from the signal intensity data Ps. Thus, the difference H is calculated for each frequency band.

[0097] ​Further, the representative data can also be constituted by the average values of the signal intensities L of the respective frequency bands fl to fn at the times Tl to Tn. In this case, for example, the signal intensity Ll of the frequency band fl included in the representative data is constituted by the average value of the signal intensities Ll_l to Ln_l, and the signal intensity Ln of the frequency band fn included in the representative data is constituted by the average value of the signal intensities Ll_n to Ln_n.

[0098] Next, the processor 101 determines the abnormality level of the vibration state of the pump 70 under each frequency band by comparing the difference H with the plurality of reference values Zl, Z2, Z3 for each frequency band (step S73). For example, in a case where the difference H is 0 or more and less than the reference value Zl (for example, 3 dB), the abnormality level is "0", and the vibration state is "normal". In a case where the difference H is the reference value Zl or more and less than the reference value Z2 (for example, 6 dB), the abnormality level is "1", and it is recommended to perform the state confirmation of the pump 70. In a case where the difference H is the reference value Z2 or more and less than the reference value Z3 (for example, 10 dB), the abnormality level is "2", and it is recommended to perform the maintenance of the pump 70. In a case where the difference H is the reference value Z3 or more, the abnormality level is "3", and it is a dangerous state in which the replacement or the like of the pump 70 is required.

[0099] The processor 101 outputs an abnormality alarm based on the determination result of the abnormality level of the vibration state in the pump 70 (step S75). Specifically, the processor 101 outputs an abnormality alarm of a high warning level (for example, "dangerous") in a case where the abnormality level is "3" (that is, in a case where H ≥ Z3), outputs an abnormality alarm of a relatively high warning level (for example, "recommended maintenance") in a case where the abnormality level is "2" (that is, in a case where Z2 ≤ H < Z3), and outputs an abnormality alarm of a low warning level (for example, "caution") in a case where the abnormality level is "1" (that is, in a case where Zl ≤ H < Z2). Further, the processor 101 can output that the vibration state of the pump 70 is "normal" in a case where the abnormality level is "0" (that is, in a case where H < Zl). Next, the processor 101 stores the determination result of the abnormality level (step S77).

[0100] The processor 101 determines a predetermined number (for example, 5) of vibration states of the pump 70 under each frequency band, in which the abnormality level is high at a reference time point (for example, the current time point), and extracts the frequency band corresponding to the determined vibration state (step S79). The reference time point is configured to be arbitrarily selected by the user.

[0101] The processor 101 performs trend prediction processing for the extracted frequency band (hereinafter, also referred to as "extracted frequency band") with respect to the vibration state (step S81). Specifically, the processor 101 predicts the trend of the difference H in the future on the basis of the time series data of the difference H in the past (past) in the extracted frequency band. For example, the processor 101 approximates the time series data of the difference H before a reference time point by an approximation curve (for example, linear approximation, exponential approximation, or the like) to predict the difference H after the reference time point. In addition, the processor 101 can also acquire a regression line by performing regression analysis on the time series data of the difference H in the past, and predict the difference H in the future on the basis of the slope and intercept of the regression line.

[0102] The processor 101 displays the result of the trend prediction processing in the form of a trend graph on the display 105 (step S83). The processor 101 stores data related to the trend graph in the memory 103 (step S85), and ends the trend analysis process. For example, the processor 101 stores various data such as the extracted frequency band, the trend graph, and the like in the memory 103.

[0103] <Screen Example>

[0104] Figure 13 is a diagram illustrating an example of the layout of the user interface screen 500. However, the user interface screen 500 can be a layout other than the layout of Figure 13 , as long as the function described later can be implemented.

[0105] Referring to Figure 13 , the user interface screen 500 includes display areas 502 to 512, a display area 514 of measurement conditions and set values, various buttons 516, display areas 520 to 540, and graphs 550 to 570.

[0106] In the display area 502, the identification number (unit number) of the sensor unit 20, the identification number (sensor number) of the sensor 30, the name of the measurement object (for example, a pump), and the like are displayed. In the display area 504, the state based on the distance MD1 calculated by the MT method using multi-dimensional data is displayed. The state changes in correspondence with the value of the calculated distance MD1. For example, the state "danger" is displayed in a case where the distance MD1 of threshold value Th1 (for example, 5) or more is calculated, the state "recommended maintenance" is displayed in a case where the distance MD1 of less than the threshold value Th1 and threshold value Th1a (for example, 4) or more is calculated, the state "caution" is displayed in a case where the distance MD1 of less than the threshold value Th1a and threshold value Th1b (for example, 3) or more is calculated, and the state "normal" is displayed in a case where the distance MD1 of less than the threshold value Th1b is calculated. In this way, the larger the distance MD1, the higher the warning level of the state.

[0107] In display area 506, a status is displayed based on the distance MD2 calculated using the MT method with two-dimensional data. This status changes accordingly with the value of the distance MD2. For example, if the calculated distance MD2 is above a threshold Th2 (e.g., 5), the status "Danger" is displayed; if the calculated distance MD2 is below the threshold Th2 but above the threshold Th2a (e.g., 4), the status "Recommended Maintenance" is displayed; if the calculated distance MD2 is below the threshold Th2a but above the threshold Th2b (e.g., 3), the status "Caution" is displayed; and if the calculated distance MD2 is below the threshold Th2b, the status "Normal". Thus, the larger the distance MD2, the higher the warning level of the status.

[0108] In display area 508, the status based on trend analysis is displayed. This status changes accordingly with the difference H at the reference time point. For example, if the difference H at the reference time point is greater than or equal to the reference value Z3 (e.g., 10 dB), the status "Danger" is displayed; if the difference H is less than the reference value Z3 but greater than or equal to the reference value Z2 (e.g., 6 dB), the status "Recommended Maintenance" is displayed; if the difference H is less than the reference value Z2 but greater than or equal to the reference value Z1 (e.g., 3 dB), the status "Caution" is displayed; and if the difference H is less than the reference value Z1, the status "Normal" is displayed. Thus, the larger the difference H, the higher the warning level of the status.

[0109] Display area 520 shows the time-series sensor data (raw data) detected by sensor 30. Display area 530 shows the analytical results (spectrum) obtained by performing FFT (fast Fourier transform) analysis on the time-series sensor data. Display area 540 shows the signal strength data obtained by performing 1 / 3 octave band analysis on the time-series sensor data using a bar chart.

[0110] Graph 550 shows the time series data at distance from MD1. Graph 560 shows the time series data at distance from MD2. Graph 570 is obtained through... Figure 12 The trend curve obtained by processing step S83.

[0111] As described above, it is predicted that pump 70 will malfunction several days after the distance MD1 above the initially calculated threshold Th1 is determined, and it is predicted that pump 70 will malfunction several hours after the distance MD2 above the initially calculated threshold Th2 is determined. Therefore, the user can infer the timing of the malfunction of pump 70 by checking curves 550 and 560.

[0112] For example, the user can make preparations for maintenance in advance while confirming the state change in the case where the distance MD1 is confirmed to be equal to or greater than the threshold Thl in the graph 550, and immediately start maintenance in the case where the distance MD2 is confirmed to be equal to or greater than the threshold Th2 in the graph 560. Thus, the timing of the equipment maintenance such as inspection, maintenance, repair, and the like of the pump 70 can be predicted with high accuracy, and thus the equipment maintenance can be implemented in a planned manner.

[0113] <FUNCTIONAL STRUCTURE>

[0114] Figure 14 is a functional block diagram of the analysis device 10. Referring to Figure 14 , the analysis device 10 includes a signal input section 202, an intensity calculation section 204, a first distance calculation section 206, a center-of-gravity calculation section 208, a second distance calculation section 210, an abnormality prediction section 212, a trend analysis section 214, and an output control section 216 as main functional structures. These functions are realized, for example, by the processor 101 of the analysis device 10 executing a program saved in the memory 103. In addition, part or all of these functions can also be realized by hardware.

[0115] The signal input section 202 accepts input of the vibration signal detected by the sensor 30 mounted to the pump 70 in operation. Specifically, the signal input section 202 receives the vibration signal (digital signal) detected by the sensor 30 via the sensor unit 20.

[0116] The intensity calculation section 204 calculates a plurality of signal intensities corresponding to a plurality of frequency bands, respectively, by analyzing the vibration signal accepted by the signal input section 202. Specifically, the intensity calculation section 204 calculates the signal intensity of each frequency band (for example, the signal intensities L of each frequency band fl to fm) by performing octave analysis (for example, 1 / 3 octave analysis) on the vibration signal corresponding to the pump 70. In addition, the intensity calculation section 204 can also be a structure that calculates the signal intensity of each frequency band by Fast Fourier Transform (FFT).

[0117] The first distance calculating section 206 calculates Mahalanobis distance (e.g., distance MD1) of a signal space (e.g., signal space X1) constituted by a plurality of signal strengths with respect to a unit space (e.g., unit space U1) set in advance using the MT method. Typically, the unit space U1 is constituted by a plurality of signal strengths L corresponding to a plurality of frequency bands f1~fn calculated by analyzing a vibration signal corresponding to the pump 70 at a normal time. The vibration signal corresponding to the pump 70 at the normal time can be, for example, a vibration signal indicating a vibration state of the pump 70 at each time T1~Tn (i.e., during the initial period of the operation start) or a vibration signal indicating a vibration state of a pump of the same kind as the pump 70 in a normal state. By calculating the distance MD1 with respect to each time (e.g., times Tn+1~Tm), time series data of the distance MD1 is generated.

[0118] The center-of-gravity calculating section 208 calculates two-dimensional center-of-gravity data (e.g., center-of-gravity positions Gx, Gy) indicating a center-of-gravity position of a plurality of signal strengths calculated by the strength calculating section 204.

[0119] The second distance calculating section 210 calculates Mahalanobis distance (e.g., distance MD2) of a signal space (e.g., signal space X2) constituted by two-dimensional center-of-gravity data with respect to a unit space (e.g., unit space U2) set in advance using the MT method. Typically, the unit space U2 is constituted by two-dimensional center-of-gravity data indicating a center-of-gravity position of a plurality of signal strengths L constituting the unit space U1. By calculating the distance MD2 with respect to each time (e.g., times Tn+1~Tm), time series data of the distance MD2 is generated.

[0120] The abnormality predicting section 212 predicts an abnormality occurrence period in which the pump 70 is abnormal based on the distance MD1 and the distance MD2. In some aspects, the abnormality occurrence period in which the distance MD1 of the threshold value Th1 or more is calculated is predicted to be in the near future than the abnormality occurrence period in which the distance MD1 of the threshold value Th1 or more is not calculated. Specifically, the abnormality predicting section 212 predicts that the pump 70 is abnormal several days after the distance MD1 of the threshold value Th1 or more is calculated. In addition, the abnormality predicting section 212 can predict that the possibility that the pump 70 is abnormal within several days from the current time point is low in a case where the distance MD1 of the threshold value Th1 or more is not calculated at the current time point.

[0121] On the other hand, the timing of the occurrence of the anomaly in the case where the distance MD1 above the threshold Thl is calculated and the distance MD2 above the threshold Th2 is calculated is predicted to be in the near future from the timing of the occurrence of the anomaly in the case where the distance MD1 above the threshold Thl is calculated and the distance MD2 above the threshold Th2 is not calculated. Specifically, the anomaly prediction unit 212 predicts that the pump 70 will occur an anomaly several hours after the distance MD2 above the threshold Th2 is calculated. Further, in the case where the distance MD1 above the threshold Thl is calculated at the current time point and the distance MD2 above the threshold Th2 is not calculated, the anomaly prediction unit 212 can predict that the probability of the pump 70 occurring an anomaly within several hours from the current time point is low.

[0122] The trend analysis unit 214 calculates, for each of the plurality of frequency bands, a difference H of the signal intensity of the frequency band corresponding to the pump 70 from the signal intensity of the frequency band corresponding to the reference pump. The trend analysis unit 214 determines the abnormality level of the vibration state of the pump 70 under each frequency band on the basis of the signal intensity of each frequency band corresponding to the pump 70 and a prescribed reference value. Specifically, the trend analysis unit 214 compares, for each of the plurality of frequency bands, the difference H under the frequency band with a plurality of reference values Zl to Z3, thereby determining the abnormality level of the vibration state of the pump 70 under the frequency band.

[0123] Further, the trend analysis unit 214 determines, from the vibration states of the pump 70 under each frequency band, a prescribed number (for example, 5) of vibration states from the side of the high abnormality level, and extracts the frequency band corresponding to the determined vibration states. The trend analysis unit 214 predicts the future difference H of each frequency band on the basis of the time series data of the difference H of each frequency band stored in the storage 103. Specifically, the trend analysis unit 214 predicts the future difference H by performing regression analysis on the time series data of the difference H before the reference time point. Alternatively, the trend analysis unit 214 predicts the future difference H by approximating the time series data of the past difference H with an approximation curve.

[0124] The output control unit 216 outputs various information such as the prediction result of the anomaly prediction unit 212. In a certain aspect, in the case where the distance MD1 above the threshold Thl is calculated, the output control unit 216 outputs first warning information (for example, information for warning that the pump 70 will occur an anomaly in the near future), and in the case where the distance MD2 above the threshold Th2 is calculated, the output control unit 216 outputs second warning information (for example, information for warning that the pump 70 will occur an anomaly in the very near future) of a warning level higher (strongly urging the user to be on guard) than the warning level of the first warning information.

[0125] For example, in a case where the first warning information is a warning level indicating a medium degree of urgency, the second warning information is a warning level indicating a high degree of urgency. That is, the second warning information is information of a higher degree of urgency than that of the first warning information. In this case, the output control section 216 can also display the second warning information in the display 105 more emphasized than the display manner of the first warning information (that is, in a manner more conspicuous than the first warning information).

[0126] In another aspect, the output control section 216 causes the display 105 to display various information shown in the user interface screen 500. Specifically, the output control section 216 causes the display 105 to display a graph 550 of time series data of the distance MD1 (for example, in Figure 13 Figure 13 In yet another aspect, the output control section 216 causes the display 105 to display a future difference H of each frequency band predicted (for example, a trend graph shown in a graph 570). In addition, the output control section 216 outputs (for example, displays in the display 105) an abnormality alarm based on the determination result of the abnormality level determined by the trend analysis section 214.

[0127] <Advantages>

[0128] According to the present embodiment, it is possible to predict, by confirming the distance MD1 and the distance MD2 based on the MT method, whether the abnormality occurrence period of the pump 70 is a near future (for example, several days later) or an extremely near future (for example, several hours later). Also, it is possible to predict the abnormality of the pump 70 in advance from various angles by referring to the trend based on the trend graph together. Therefore, it is possible to implement equipment maintenance in a planned manner.

[0129] <Other Embodiments>

[0130] (1) In the above-described embodiment, a structure in which the analysis device 10 predicts that the abnormality occurs several days later from when the Mahalanobis distance MD1 of the threshold value Th1 or more is first calculated is described. With regard to this, as shown in Figure 9 the Mahalanobis distance MDx1 is the threshold value Th1 or more at times Ta2, Ta3, and Ta4, and it is referred to that the pump has an abnormality at time Ta5. Therefore, it can be said that the more the number of times that the Mahalanobis distance MDx1 is the threshold value Th1 or more, the closer the abnormality occurrence period. Therefore, it can also be that the more the number of times that the distance MD1 of the threshold value Th1 or more is calculated, the closer the future in which the analysis device 10 (abnormality prediction section 212) predicts that the pump 70 has an abnormality.

[0131] ​(2) In the above-described embodiments, a program that causes a computer to function to perform control as described in the above-described flowcharts can also be provided. Such a program can also be recorded in a non-transitory computer-readable recording medium such as a floppy disk, a CD-ROM (Compact Disk Read Only Memory), a secondary storage device, a main storage device, a memory card, and the like that is attached to a computer, to be provided as a program product. Alternatively, the program can also be provided by being recorded in a recording medium such as a hard disk built in a computer. Further, the program can also be provided by being downloaded via a network.

[0132] The program can also be used to call necessary modules among program modules provided as a part of an operating system (OS) of a computer in a prescribed order at a prescribed timing to execute processing. In this case, the program itself does not include the above-described modules but cooperates with the OS to execute processing. Such a program that does not include the modules can also be included in the program according to the present embodiment. Further, the program according to the present embodiment can also be provided as a part of another program. In this case as well, the program itself does not include modules included in the above-described another program but cooperates with the another program to execute processing. Such a program that is embedded in another program can also be included in the program according to the present embodiment.

[0133] (3) The structure exemplified as the above-described embodiments is an example of the structure of the present application, can also be combined with other known technologies, and can also be changed in a manner such as omission of a part within a range not departing from the gist of the present application. Further, it can also be the case that, in the above-described embodiments, the processing and the structure described in other embodiments are appropriately adopted to be implemented.

[0134] It is to be understood that the embodiments disclosed this time are by way of illustration and not by way of limitation. The scope of the present application is not shown by the above-described description but shown by the claims, and intended to include all modifications equivalent in meaning and scope to the claims.

[0135] Explanation of Reference Numerals

[0136] 10: analysis device; 20: sensor unit; 21: filter; 22: amplifier; 23: converter; 30: sensor; 40: terminal device; 50: network; 70: pump; 100: vibration analysis system; 101: processor; 103: memory; 105: display; 107: input device; 109: signal input interface; 111: communication interface; 202: signal input section; 204: intensity calculation section; 206: first distance calculation section; 208: center of gravity calculation section; 210: second distance calculation section; 212: anomaly prediction section; 214: trend analysis section; 216: output control section; 310, 320: data set; 500: user interface screen; 1000: system.

Claims

1. A vibration analysis system comprising: a signal input section that accepts input of a vibration signal detected by a sensor mounted to an object in motion; a strength calculation section that calculates a plurality of signal strengths corresponding to a plurality of frequency bands by analyzing a vibration signal corresponding to the object; a first distance calculation section that calculates a first Mahalanobis distance of a first signal space constituted by the plurality of signal strengths with respect to a first unit space set in advance; a center-of-gravity calculation section that calculates two-dimensional center-of-gravity data representing a center-of-gravity position of the plurality of signal strengths calculated by the strength calculation section; a second distance calculation section that calculates a second Mahalanobis distance of a second signal space constituted by the two-dimensional center-of-gravity data with respect to a second unit space set in advance; and an abnormality prediction section that predicts an abnormality occurrence timing at which the object is abnormal based on the first Mahalanobis distance and the second Mahalanobis distance.

2. The vibration analysis system according to claim 1, wherein the first unit space is constituted by a plurality of signal strengths corresponding to the plurality of frequency bands calculated by analyzing a vibration signal corresponding to the object at a normal time, and the second unit space is constituted by two-dimensional center-of-gravity data representing a center-of-gravity position of the plurality of signal strengths constituting the first unit space.

3. The vibration analysis system according to claim 1 or 2, wherein the abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than a first threshold value is predicted to be in the near future than the abnormality occurrence timing in a case where the first Mahalanobis distance is not calculated to be equal to or higher than the first threshold value.

4. The vibration analysis system according to claim 3, wherein the abnormality prediction section predicts that the object is abnormal several days after the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value.

5. The vibration analysis system according to claim 3, wherein the abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value and the second Mahalanobis distance is calculated to be equal to or higher than a second threshold value is predicted to be in the near future than the abnormality occurrence timing in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value and the second Mahalanobis distance is not calculated to be equal to or higher than the second threshold value.

6. The vibration analysis system according to claim 5, wherein the abnormality prediction section predicts that the object is abnormal several hours after the second Mahalanobis distance is calculated to be equal to or higher than the second threshold value.

7. The vibration analysis system according to claim 5, further comprising an output control section that outputs first warning information in a case where the first Mahalanobis distance is calculated to be equal to or higher than the first threshold value, and outputs second warning information of which a warning level is higher than a warning level of the first warning information in a case where the second Mahalanobis distance is calculated to be equal to or higher than the second threshold value.

8. The vibration analysis system according to claim 7, wherein the output control section causes a display to display time series data of the first Mahalanobis distance and time series data of the second Mahalanobis distance. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ 9. A vibration analysis method comprising the steps of: accepting input of a vibration signal detected by a sensor mounted on an object in motion; calculating a plurality of signal intensities corresponding to a plurality of frequency bands, respectively, by analyzing the vibration signal corresponding to the object; calculating a first Mahalanobis distance of a first signal space constituted by the plurality of signal intensities with respect to a first unit space set in advance; calculating barycentric data of two dimensions indicating a position of a barycenter of the plurality of signal intensities calculated; calculating a second Mahalanobis distance of a second signal space constituted by the barycentric data of two dimensions with respect to a second unit space set in advance; and predicting an abnormality occurrence period in which the object is abnormal based on the first Mahalanobis distance and the second Mahalanobis distance. 9.

Images